Youth unemployment and inactivity - Norden

TemaNord 2015:548

Youth unemployment and inactivity

TemaNord 2015:548

Ved Stranden 18 DK-1061 Copenhagen K www.norden.org

Youth unemployment and inactivity A comparison of school-to-work transitions and labour market outcomes in four Nordic countries

Young people follow highly different trajectories from age 16 up to age 20, a time period which is often argued to be the most critical in terms of their future labour market outcomes. The focus of this report is on investigating the look of these early pathways, as well as on exploring their link to labour market outcomes in adulthood. Results are reported and compared for four Nordic countries: Denmark, Finland, Norway and Sweden.

TemaNord 2015:548 ISBN 978-92-893-4229-2 (PRINT) ISBN 978-92-893-4230-8 (PDF) ISBN 978-92-893-4231-5 (EPUB) ISSN 0908-6692

TN2015548 omslag.indd 1

20-07-2015 12:48:57

Youth unemployment and inactivity A comparison of school‐to‐work transitions and labour market outcomes in four Nordic countries Karsten Albæk, Rita Asplund, Erling Barth, Lena Lindahl, Kristine von Simson and Pekka Vanhala

TemaNord 2015:548

Youth unemployment and inactivity A comparison of school‐to‐work transitions and labour market outcomes in four Nordic countries Karsten Albæk, Rita Asplund, Erling Barth, Lena Lindahl, Kristine von Simson and Pekka Vanhala ISBN 978‐92‐893‐4229‐2 (PRINT) ISBN 978‐92‐893‐4230‐8 (PDF) ISBN 978‐92‐893‐4231‐5 (EPUB) http://dx.doi.org/10.6027/TN2015‐548 TemaNord 2015:548 ISSN 0908‐6692 © Nordic Council of Ministers 2015 Layout: Hanne Lebech Cover photo: ImageSelect Print: Rosendahls‐Schultz Grafisk Printed in Denmark

This publication has been published with financial support by the Nordic Council of Ministers. However, the contents of this publication do not necessarily reflect the views, policies or recom‐ mendations of the Nordic Council of Ministers.

www.norden.org/nordpub

Nordic co‐operation Nordic co‐operation is one of the world’s most extensive forms of regional collaboration, involv‐ ing Denmark, Finland, Iceland, Norway, Sweden, and the Faroe Islands, Greenland, and Åland. Nordic co‐operation has firm traditions in politics, the economy, and culture. It plays an im‐ portant role in European and international collaboration, and aims at creating a strong Nordic community in a strong Europe. Nordic co‐operation seeks to safeguard Nordic and regional interests and principles in the global community. Common Nordic values help the region solidify its position as one of the world’s most innovative and competitive. Nordic Council of Ministers Ved Stranden 18 DK‐1061 Copenhagen K Phone (+45) 3396 0200 www.norden.org

Contents Introduction ................................................................................................................................................. 7 Increasing focus on school‐to‐work transitions .................................................................... 7 Key features of the study............................................................................................................... 10 Structure and content of the present report ......................................................................... 12 1. Performance of young people – an international perspective ...................................... 17 1.1 Setting the stage .................................................................................................................. 17 1.2 Snapshot of recent Nordic comparative studies .................................................... 20 1.3 The Nordic story told by international comparative indicators ..................... 21 2. General description of post‐compulsory‐school activities ............................................. 37 2.1 The national datasets used ............................................................................................. 37 2.2 Main activities of young people – the overall pattern ......................................... 41 2.3 Non‐completers – an introductory view ................................................................... 65 2.4 Individual school‐to‐work profiles – a general picture ...................................... 76 3. School‐to‐work trajectories: country‐specific cluster results ....................................... 89 3.1 The cluster analysis method – a brief outline ......................................................... 89 3.2 Country‐specific cluster results: all young people ............................................... 98 3.3 Country‐specific cluster results: non‐completers .............................................. 113 4. School‐to‐work trajectories: common Nordic cluster results .................................... 129 4.1 Common Nordic trajectories and clusters: all young people ........................ 130 4.2 Common Nordic trajectories and clusters: non‐completers ......................... 136 5. Non‐completers’ school‐to‐work trajectories: stylized cluster results .................. 147 6. Labour market outcomes as young adults .......................................................................... 171 6.1 Main activities beyond age 21 – all young people ............................................. 171 6.2 Main activities beyond age 21 – young men vs. young women .................... 181 6.3 Main activities beyond age 21 – completers vs. non‐completers ................ 190 6.4 Main activities beyond age 21 – late completion vs. non‐completion ....... 210 7. Labour market outcomes in view of background ............................................................ 223 7.1 Gender and labour market outcomes ..................................................................... 223 7.2 Cohort and labour market outcomes ...................................................................... 227 7.3 Family background and labour market outcomes ............................................. 232 7.4 Early pathways through education and labour market outcomes ............. 241 8. Summary and discussion ........................................................................................................... 273 References ................................................................................................................................................ 279 Sammanfattning .................................................................................................................................... 285 Appendix: Descriptions of national datasets used .................................................................. 293 Denmark ............................................................................................................................................ 293 Finland ............................................................................................................................................... 294 Norway .............................................................................................................................................. 295 Sweden............................................................................................................................................... 296

Introduction Youth unemployment has for several decades been a much debated topic in both the academic and the political arena. The recent global economic crisis, in combination with profound changes in labour market structures (see e.g. Acemoglu and Autor, 2011), shifted youth unemployment to the top of the political agenda after alarming reports on surging youth unem‐ ployment rates and growing risks of young people becoming economically and socially marginalised (e.g. Scarpetta et al., 2010; ILO, 2011, 2013; Räisänen, 2013). Concomitantly, increasing attention was paid to so‐called NEETs, that is, young people not in employment, education or training. Several recent reports highlight the prevalence of this phenomenon and also provide estimates of the societal costs of early school leaving and NEETs (e.g. Eurofound, 2012; Brunello and De Paola, 2014). Other studies highlight the complexity of problems often associated with being a NEET. A survey of the Prince Trust (2010) covering young people aged 16–25 reports that NEETs are significantly more likely to feel ashamed, rejected, lost, anxious, insecure, down and depressed, isolated and unloved. They are also disproportionately more likely to say that they had turned to drugs and that their life had no direction.

Increasing focus on school‐to‐work transitions A growing body of evidence identifies early school leaving and school dropout as major reasons underlying youth unemployment and margin‐ alisation from working life in a world where an upper secondary certifi‐ cate has increasingly become the minimum requirement for proper ac‐ cess to the job market. Those left behind at the lowest floor – the low‐ or unskilled school leavers lacking work experience – are the losers in the competition with the better skilled. Their disproportionate presence among those holding temporary jobs, in combination with their high concentration in cyclically sensitive industries, makes them especially vulnerable. Apart from a much higher risk of becoming unemployed, as compared to their higher‐educated peers, young people with a low edu‐ cation also face a much higher risk of repeated spells of unemployment

that, moreover, tend to prolong over time. Ultimately school leavers end up outside the labour force for shorter or longer periods of time. Hence, low educational attainment levels not only impede initial in‐ sertion into the labour market, but also represent an enduring barrier to employment (e.g. OECD, 2008a). Indeed, while the negative labour mar‐ ket and social outcomes of youth unemployment of a more temporary nature have been shown to diminish over time, there is a considerable body of evidence on long‐term unemployment having lasting scarring effects on young people in terms of both future employment and future wages. Illustrative examples are a study by Kahn (2010) for the USA showing that graduating from college in a bad economy has large, nega‐ tive and persistent effects on wages, and a study by Bell and Blanchflow‐ er (2009) for the UK reporting that youth unemployment raises unem‐ ployment, lowers wages, worsens health and lowers job satisfaction still 25 years later. Studies reporting permanent employment and wage loss‐ es are also found for other European countries, including Germany (e.g. Schmillen and Umkehrer, 2013; Möller and Umkehrer, 2014), Norway (e.g. Nilsen and Holm Reiso, 2014) and Sweden (e.g. Nordström Skans, 2011). Taken together, the existing evidence implies that the long‐term costs of early exit from the education system have increased. While the economic business cycle tends to have long‐lasting career consequences for young labour market entrants, also the schooling deci‐ sion itself may be affected by the state of the labour market. Desire to work may be an important motivation for some youth to end their edu‐ cation at an earlier stage. According to standard human capital theory (Becker, 1964), the economic situation affects the schooling decision mainly through the opportunity cost of schooling, measured as foregone earnings. A booming labour market implies more job openings and high‐ er wages, which in turn increase the opportunity cost of schooling. This may encourage some youth to take advantage of the favourable labour market conditions and leave school on a temporary or permanent basis. An economic downturn, on the other hand, may induce youth to stay in school and postpone their labour market entry, or to return to educa‐ tion. Previous research on upper secondary schooling supports this pre‐ diction: Students are less likely to stay in school in good times. Also en‐ rolment tends to decrease in good times (Clark, 2011), with graduation rates being countercyclical, as well (Reiling and Strøm, 2015). Addition‐ ally, the propensity to interrupt the education and drop out from school altogether increases in economic upturns (von Simson, 2014). On the whole, though, the association is not very strong and other factors, such

8


as parental education, are more important in explaining schooling deci‐ sions than is the economic situation. Nonetheless, early school leaving and dropout from upper secondary education can be seen to typically represent unsuccessful transitions within the school system and further into working life. This contention has, more recently, broadened the focus on unemployment and margin‐ alisation among young people to the role of school‐to‐work transitions. Indeed, the transition from education to working life has frequently been argued to be the most critical phase in terms of a young person’s labour market outcomes later in life. While these transitions evidently reflect the family situation and compulsory‐school outcomes, these much researched background factors cannot offer a full explanation for young people’s highly different school‐to‐work‐transition patterns and, ultimately, for their differences in labour market outcomes in adulthood. The transitions in themselves seem to be of importance, as well. In other words, it seems reasonable to assume that young people’s early school and labour market experiences upon leaving compulsory school also play a role for their labour market outcomes later in life. These contentions explain and motivate the focus of our study on four Nordic countries – Denmark, Finland, Norway and Sweden – the main results of which are reported in this volume. More precisely, first we present and discuss the school‐to‐work transitions of young people leaving compulsory school. Next we use this information to indicate the extent to which different school‐to‐work‐transition patterns are related to the probability of young people ending up in alternative labour mar‐ ket situations in adulthood. Doing so, we also include basic information on family background, which allows us to test whether or not early post‐ compulsory‐school experiences have a “signal value of their own” espe‐ cially when it comes to young people’s probability of going into risky labour‐market tracks dominated by time spent in NEET activities, that is, in unemployment or entirely outside the labour force. Put differently, is it evident that, in addition to family background, also early school‐to‐ work transitions contain important information concerning the likeli‐ hood of different labour market outcomes later in life? The rest of this chapter provides a more detailed outline of our study and the structure of the present report.


9

Key features of the study The labour market performance of youth is commonly described by use of various rather conventional indicators highlighting phenomena such as early school leaving, premature training dropout, youth employment and unemployment, and NEET. These indicators share the feature of providing an instantaneous picture of the educational outcome or of the position/status of youth in the labour market. The mediated picture can, as a consequence, be partly misleading as these static indicators fail to see the observed outcome “as a cumulative process of disengagement or withdrawal that occurs over time” (Lyche, 2010, p. 14). Emerging indica‐ tors, providing a more dynamic description of youth labour market per‐ formance, can correct for some of the shortcomings characterising static indicators but share, nonetheless, the common drawback of oversimpli‐ fying the dynamic nature of youth school‐to‐work trajectories. In order to overcome this unsatisfactory situation, growing numbers of researchers have used alternative ways to assess comprehensively the multiple patterns of school‐to‐work transitions that youth are known to face when moving from education into working life. However, these tran‐ sitions from education to work are far from easy to measure because of the “fluidity of the youth labour market” (OECD 2008b, p. 59). In particu‐ lar, school‐to‐work transitions are often long‐lasting processes involving frequent status changes between education, temporary jobs, unemploy‐ ment and inactivity (Müller and Gangl, 2003; Wolbers, 2007; Saar et al., 2008). This also explains why transitions from education to work cannot be fully understood by analysing single changes of status only. More recent research on school‐to‐work transitions is therefore in‐ creasingly based on longitudinal data allowing young people to be fol‐ lowed after they have completed their education over a longer time pe‐ riod. However, apart from high‐quality longitudinal data, such a dynamic approach also requires the use of proper statistical methods. Since young people often shift between education, inactivity, unemployment and work before getting a stable job, sequence and cluster analysis has often been argued to be the most appropriate methodological approach in this context (e.g. Brzinsky‐Fay, 2007). A strong advantage of this soci‐ ological method is that it allows entire individual school‐to‐work trajec‐ tories, including nature of spells and their order, to be identified, com‐ pared and classified into one of several distinct types of youth trajecto‐ ries. This categorisation also allows proper account to be made for the fact that youth trajectories often show a high degree of diversity espe‐ cially by gender and educational background. Accordingly, this evidence

10


clearly improves our understanding of young people’s often bumpy, occasionally unsuccessful, transition into employment. As already noted, this dynamic approach to analysing school‐to‐work trajectories requires access to high‐quality longitudinal data. It is, there‐ fore, hardly surprising that this kind of evidence is, so far, available only for a small number of countries, occasionally for the whole of Europe as averages using information on a limited number of European countries, as done in Quintini and Manfredi (2009). Also missing is broad‐based comparative information for individual Nordic countries, a knowledge gap that the present report aims to fill. Indeed, this report is, to our knowledge, the first contribution using sequence and cluster analysis to investigate school‐to‐work transitions for full cohorts of young people and, moreover, to also undertake cross‐country comparisons of such transitions among young people based on comparative national data. In exploring the educational and labour market experiences of young people, we pay particular attention to identifying risky trajectories and conspicuous differences in these respects between youth differing in their educational background, while also differentiating across genders. Additionally, attempts are made to identify and quantify possible chang‐ es in early transition patterns within countries by comparing key out‐ comes for a total of three youth cohorts: 16‐year‐olds in 1993, 1998 and 2003, respectively. All three cohorts are traced up to the year 2008 im‐ plying that the shortest follow‐up period covers 5 years and the longest 15 years. Equally important, school‐to‐work‐transition patterns and changes in these patterns are compared across the four Nordic countries under study to highlight major similarities and dissimilarities in “typi‐ cal” school‐to‐work trajectories. The analysis briefly outlined above provides detailed cross‐Nordic in‐ formation on distinct types of school‐to‐work trajectories for our three cohorts of 16‐year‐olds, up to age 20. Finally, this information is linked to the young persons’ labour market outcomes at three points later in time: when they turn 21, 26 and 31. Throughout, these alternative la‐ bour market activities are grouped into five broad categories – full‐time studying, employment, unemployment, disability benefits and “other” (inactivity). Apart from simply describing these relations between early and later labour market outcomes, we also undertake statistical analysis by use of so‐called multinomial logit techniques, with a view of high‐ lighting the relative strength between different early school‐to‐work‐ transition patterns and alternative labour market outcomes later as a young adult. As explained earlier, we thereby also account for family background in order to explore whether the link observed between ear‐


11

ly school‐to‐work‐transition patterns and alternative labour market outcomes in adulthood remains approximately unchanged, diminishes or eventually disappears when adding parental information. The last option would, of course, mean that the distinct types of early school‐to‐ work trajectories identified merely reflect differences in the young per‐ sons’ family background. Likewise, the first option would point to the opposite: no link whatsoever between early school‐to‐work trajectories and family background. Taken together, all this new evidence for the four Nordic countries under study should provide policy‐makers valua‐ ble information to guide them in their decisions on actions aimed to enhance the transition of youth from school into working life.

Structure and content of the present report The report is structured as follows. The next chapter provides a snap‐ shot of what widely used international indicators can tell us about edu‐ cational outcomes, school‐to‐work transitions and labour market per‐ formance of young people in the Nordic countries. Simultaneously, the information mediated by these mainly static indicators serves as a benchmark when we in the next chapters turn to presenting results ob‐ tained based on our national longitudinal data. Chapter 2 introduces the national datasets underlying the empirical evidence reported in this volume. It also describes and compares across countries the main post‐compulsory‐school activities in which the young people covered by our data have been engaged in, with the emphasis being on their situation up to age 20. In other words, this chapter paints a general picture of young people’s activities over a time period which is often argued to be the most critical in terms of their future labour mar‐ ket outcomes. The chapter also introduces our preferred definition of young people lacking an upper secondary certificate: these young people are called “non‐completers”, as compared to “completers”, implying that they only have their compulsory‐school exam still five years after having left primary education. In the next chapter, Chapter 3, we present a first set of results ob‐ tained by “grouping” the multitude of individual school‐to‐work‐ transition pathways followed by young people when aged 16 to 20, as reported in Chapter 2, by use of so‐called cluster analysis. Accordingly, the chapter starts with a brief outline and discussion of the basic idea of the cluster analysis method. Only then we turn to the cluster results produced for each country using information on the full youth popula‐

12


tion. For comparative purposes, separate results are presented also for the non‐completers. All these results on “typical” early school‐to‐work‐ transition patterns are obtained by allowing each national dataset to form the clusters for the country in question, for which reason we refer to them as “country‐specific cluster results”. In other words, we do not restrict the cluster analysis in order to produce as similar clusters as possible across the four Nordic countries under study. The country‐specific cluster results reported in Chapter 3 are used as key inputs for the cluster analyses undertaken in the next two chapters, where we explicitly aim at improving the comparability of early school‐ to‐work trajectories across the four Nordic countries by identifying what we have labelled “common” Nordic school‐to‐work trajectories (Chapter 4) and “stylized” school‐to‐work pathways for the Nordic countries (Chapter 5). For the cluster analysis undertaken in Chapter 4 we have for each country made a list of the observed trajectories (sequences of activities) from age 16 up to age 20 and then calculated the number of young people following each of these sequences. This country‐specific information is then pooled into one big data to which we apply cluster analysis in order to allocate all these trajectories of Nordic youth across ten clusters “common” for the four Nordic countries under study.1 This approach allows interesting patterns common to the Nordic countries to be identified. Also this analysis is undertaken separately for all young people and non‐completers. In Chapter 5, the focus is on comparing a set of “stylized” school‐to‐ work pathways constructed for the four Nordic countries under study. In brief, the basic idea is to first select a number of “typical” early school‐ to‐work trajectories based on the results obtained in Chapter 3, and then allocate the trajectories of all other young people across these “typical” pathways. By using identical “typical” early school‐to‐work trajectories as the point of departure for all four Nordic countries, we are able to shed further light on both similarities and dissimilarities across the four countries in relation to young people’s post‐compulsory‐school experi‐ ences up to age 20. However, in contrast to the analyses reported in the previous chapters, Chapter 5 is restricted to non‐completers only, that is,

────────────────────────── 1 Hence, for producing the common Nordic cluster results reported in Chapter 4, we have not merged indi‐ vidual‐level data for the four Nordic countries under study. Instead, we have merely combined information on individual trajectories showing sequences of activities from age 16 up to age 20. We have asked for and also received permission from the respective statistical bureaus to undertake such a pooling of country‐ specific individual trajectories.


13

to those young people who have no degree beyond the compulsory‐ school exam still five years after leaving basic education. A main reason for this choice of focus is that the group of young completers is highly similar in the four countries, whereas we see both striking similarities and distinct differences between the four Nordic countries when it comes to non‐completers. By complementing the results produced for non‐completers in the previous chapters with information provided by these “stylized” school‐to‐work pathways, we obtain a fuller picture of the multitude of early post‐compulsory trajectories followed by Nordic non‐completers. After Chapter 5, our focus turns from exploring young people’s main activities and school‐to‐work‐transition patterns over the five years following upon compulsory education (i.e., from age 16 up to age 20) to investigating what happens to these youngsters after they have turned 20. What kind of main activities – studying, employment, unemploy‐ ment, disability arrangements or other types of inactivity – are they mostly engaged in as young adults? Can we observe distinct and rather stable differences in this respect across genders, on the one hand, and between those differing in their educational background, on the other hand? Or is it possible that these later outcomes are, by and large, quite similar for young men and women, as well as for early and later com‐ pleters of an upper secondary degree and, possibly, even for adult non‐ completers, i.e. those with no exam beyond primary education still in adulthood? Can we identify conspicuous variations across cohorts obvi‐ ously related to fluctuations in the economic environment, or is the eventual impact of economic shocks rather outweighed by other pro‐ cesses and mechanisms affecting the labour market situation of young people representing different cohorts? Last, but not least, can we identi‐ fy clear‐cut similarities or dissimilarities in all these important dimen‐ sions across the four Nordic countries under study? Chapter 6 sets out to provide answers to these key questions. While Chapter 6 gives a description of young people’s labour mar‐ ket situation in adulthood, Chapter 7 looks into the background of these young people in an attempt to identify factors that seem to be especially strongly related to the labour market outcomes observed up to age 31. Our statistical analysis relies on so‐called multinomial logit models which show the probability of belonging to one of several mu‐ tually exclusive groups, given a particular set of background character‐ istics. In our case, these groups are made up of the five main categories of labour market activities used in the previous chapters: full‐time student, employed, unemployed, disability beneficiary or outside all of

14


these activities (“other”). The background factors accounted for in these multinomial logit models can be divided basically into two groups: one reflecting family background and the other early school‐ to‐work‐transition patterns, that is, trajectories followed straight after completion of compulsory education up to age 20. Additionally, we account for gender as well as cohort. Accounting for cohort is relevant as we base our analysis on the pooled information available for all three youth cohorts under scrutiny, i.e., those young people who turned 16 in 1993, 1998 and 2003, respectively. The last chapter of this report, Chapter 8, gathers and discusses the main findings presented in the previous chapters. While each chapter presenting our results, that is, Chapters 3 to 8, contains several sections titled “Main findings” – one for each sub‐chapter – this concluding chap‐ ter aims to draw, based on our multifaceted results, a broader picture for the four Nordic countries under study. In other words, in this concluding chapter we choose to overlook the multitude of more detailed findings reported and discussed in the separate chapters and sub‐chapters. This does by no means imply that these details are trivial. Instead, this way of presenting the main results of our study is the product of a deliberate choice with overall conclusions and remarks given in Chapter 8, with theme‐specific conclusions drawn together in the sections titled “Main findings”, and detailed country‐specific as well as cross‐country results and conclusions reported and discussed throughout the text.


15

1. Performance of young people – an international perspective The main focus of this chapter is on exploring what internationally wide‐ ly‐used indicators can tell us about educational outcomes, school‐to‐ work transitions and labour market performance of young people in the Nordic countries. Simultaneously, the information mediated by these overwhelmingly static indicators serves as an important benchmark when we, starting in the next chapter, turn to presenting results ob‐ tained based on our national longitudinal datasets.

1.1 Setting the stage The youth population has been the most affected by the recent global crisis. There are several obvious reasons why young people have been hit so hard (see e.g. OECD, 2009, 2010a; Scarpetta et al., 2010). School‐ leavers are often the first to encounter difficulties: when the labour market deteriorates, employers shed workers and also become much more selective in their hiring of new staff. As those making the transition from school to work compete with more experienced workers for (few‐ er) jobs, they often face virtually impossible barriers when trying to get a foothold in the labour market. However, the crisis has posed challeng‐ es also to those youth who were already in the labour market but hold‐ ing temporary jobs and/or working in business‐cycle sensitive indus‐ tries; they have often been among the first to lose their jobs. And with the labour market having become more selective, the risk of unemploy‐ ment for recent entrants is notably higher among those lacking relevant skills or experience and, conversely, they also face particular difficulties in finding a new job. The relatively higher vulnerability of youth to unemployment and inactivity was, in effect, a widely recognised problem in many Europe‐ an countries even before the onset of the economic crisis. Particular attention was thereby paid to the multiple barriers in finding work

faced by low‐skilled youth, that is, early school‐leavers. Indeed, many European economies were already before the crisis tackling a number of labour market problems – judged to affect adversely the transition from school to work of youth as well as their initial labour market ex‐ periences – in order to cope with unacceptably high youth unemploy‐ ment and inactivity rates. The recent economic crisis is commonly seen to have aggravated many of these structural problems and, consequently, the situation of especially those youth whose labour market prospects were weak al‐ ready prior to the crisis. The Nordic countries are no exception to this pattern, as shown in statistics compiled by notably Eurostat, ILO and OECD. In particular, although there were significant pre‐crisis differ‐ ences both in the level and evolution of youth unemployment also across the Nordic countries, they are nonetheless characterised by two distinct features which they share with the rest of Europe. First, youth face a clearly higher risk of unemployment than adults also in Northern Europe. However, while the youth/adult unemployment ratio (for 2008) falls within the interval 2 to 3 for most OECD countries, it rang‐ es between 3 and 4 in seven countries, three of which are located in the northern part of Europe (Denmark, Finland and Norway), and settles above 4 in only two countries – Iceland and Sweden (Scarpetta et al., 2010, pp. 11–12). Second, all Nordic countries have experienced a marked increase in youth unemployment since the recession began. Increasing youth unemployment rates in combination with discour‐ aging estimates of the likely short‐term evolution of youth unemploy‐ ment soon triggered, in individual countries, a multitude of actions aimed at cushioning the effects of the crisis on youth while, simultane‐ ously, pushing forward the long‐term agenda of necessary structural reforms for tackling pre‐crisis youth unemployment problems. A major challenge has thereby been to devise short‐term measures which do not conflict with but, preferably, complement and support the long‐term reform agenda of promoting more and better jobs for youth in response to projected demographic changes. The short‐ and long‐term measures planned and realised in individu‐ al countries have been surrounded by a myriad of activities initiated, not least, by the European Commission and the OECD. The many initiatives of the European Commission were brought to a head in the Europe 2020 framework launched in March 2010 [COM(2010)2020] as a continuation of the Lisbon process. This EU strategy for smart, sustainable and inclu‐ sive growth has a strong youth dimension, as have several of its accom‐ panying flagship initiatives, most notably “Youth on the Move”

18


[COM(2010)477] and “An Agenda for New Skills and Jobs” [COM(2010)682], but also the Horizon 2020 financial instrument aimed at securing Europe’s global competitiveness. Illustrative examples of recent actions supplementing the Youth on the Move education and em‐ ployment initiative include an action plan to reduce early school leaving in the EU [COM(2011)18, COM(2011)19, SEC(2011)96], and the Youth Opportunities Initiative [COM(2011)933] launched in December 2011, which can be described as a set of measures, planned for 2012 and 2013, to drive down youth unemployment. The Youth Employment Package adopted in December 2012 can be seen as a key milestone of this YOI. Most notably, this package of Commission proposals to fight youth un‐ employment recommended that member states introduce a Youth Guar‐ antee, launch a consultation of European social partners on a Quality Framework for Traineeships, and announce a European Alliance for Apprenticeships. The EU countries endorsed the principle of the Youth Guarantee in April 2013. Until the end of 2014, the EU provided an ad‐ vice service on apprenticeship and traineeship schemes in order to sup‐ port its Member States to develop high quality apprenticeship and train‐ eeship programs. Complementary to the Youth on the Move flagship initiative of the Eu‐ rope 2020 strategy is, inter alia, the EU Youth Strategy: Council Resolution on the renewed framework for European cooperation in the youth field (2010–2018) (OJ C 311, 19.12.2009, pp. 1–11). A particular feature of this strategy is that an EU Youth Report is to be drawn up at the end of each three‐year cycle to evaluate the progress made towards the overall objec‐ tives of the strategy, on the one hand, and to serve as a basis for establish‐ ing a set of priorities for the coming work cycle, on the other. The first work‐cycle EU Youth Report [COM(2012)495 final] was published in Sep‐ tember 2012 and adopted as a Joint Council–Commission Report in No‐ vember 2012. In relation to the EU Youth Strategy, a dashboard of EU Youth Indicators was released in 2011 [SEC(2011)401]. Key instruments to support the EU Youth Strategy are, most notably, the Lifelong Learning and Youth in Action programs and the Erasmus for All program. Among the many recent initiatives of the OECD, two in particular de‐ serve to be mentioned here. The first is the High‐Level Policy Forum on Jobs for Youth: Addressing Policy Challenges in OECD Countries, which was organised jointly with the Norwegian Ministry of Labour in late September 2010. The main issues and policy recommendations on how to tackle youth unemployment problems presented at this Forum, and later published in a comprehensive report (OECD, 2010b), synthesised the findings of thematic reviews of Jobs for Youth undertaken over the


19

years 2006–2009 in 16 member countries. While the review for Norway (OECD, 2008c) took place against the background of a buoyant economy, the corresponding review for Denmark (2010c) was more concerned with youth unemployment in the context of the ongoing economic crisis. These thematic reviews did not cover Finland, Iceland or Sweden, though. This High‐Level Policy Forum was preceded, about one week earlier, by a joint ILO–IMF conference – also arranged in Oslo but in co‐ operation with the Office of the Prime Minister of Norway – on The Chal‐ lenges of Growth, Employment and Social Cohesion, one focus of which was youth unemployment (ILO–IMF, 2010). A second key initiative of the OECD is the so‐called OECD Action Plan for Youth, launched in 2013 as an integral part of OECD’s work on youth, and the high‐level meetings and comprehensive reports related to the implementation of the plan (http://www.oecd.org/employment/action‐plan‐youth.htm).

1.2 Snapshot of recent Nordic comparative studies The Nordic Council of Ministers has also in recent years initiated several reports on how the youth unemployment problem is addressed in the Nordic countries. Especially the following reports should be mentioned in this context. A report edited by Markussen (2010) focuses on dropout in upper secondary education in the Nordic countries. In particular, for each Nordic country it provides an overview of the structure of upper secondary education, looks at research and results on dropouts, gives an overview of implemented measures to reduce dropout and improve through‐put of students and, finally, assesses possible effects of the im‐ plemented measures. It concludes by pointing to the need for further reforms to reduce dropout and improve levels of upper secondary com‐ pletion. Another report, by Engberg (2014), continues on this topic in the sense that it describes the reforms and other actions carried out in the Nordic countries concerning vocational and apprenticeship training, and also reflects on the challenges characterising these systems, using desk studies and interviews with national experts as analytical tools. Still another report of some relevance in this context describes both existing and planned measures, as initiated by relevant government departments, to prevent youth unemployment in the Nordic countries (Ramböll Management Consulting AB, 2010). The report reviews both short‐term measures implemented during the economic crisis and long‐ term measures related to future demographic challenges. This report

20


also relies on desk studies and interviews with national experts and public servants within the fields of education and labour markets. While this kind of comparative analysis of short‐ and long‐term measures can provide important information on what is done and planned – and in relation to which target group(s) and with which ex‐ pected outcomes in mind – in different countries, they can tell little, if anything, about the genuine impact of the measures undertaken. Such information can be obtained only by means of carefully designed and performed evaluation studies. However, the international research fo‐ cusing on in‐depth evaluation of various modes of measures directed at disadvantaged and/or unemployed youth is still surprisingly scant. More important, the evidence produced by such evaluations has, so far, been rather discouraging when it comes to both impact and effectiveness [see e.g. the reviews by Asplund (2009) and Asplund and Koistinen (2014) and, especially, the references therein]. Also worth mentioning in this context is a recent report by Halvorsen et al. (2012) on the transition between school and work for particularly vulnerable groups of youth. It concludes that between 2% and 5% of the youth cohorts are already “outsiders”. The report summarises the existing knowledge on vulnerable youth, discusses challenges and policy measures in the Nordic countries, with particular attention paid to youth, and sug‐ gests the building of a Nordic knowledge bank for good practices.

1.3 The Nordic story told by international comparative indicators Another source highlighting the youth unemployment problem from a cross‐country perspective contains indicators developed and compiled mainly by Eurostat and OECD, and published on a regular basis. Excel‐ lent examples are Eurostat’s statistical portraits of youth in Europe and the OECD’s Education at a Glance reports. In this sub‐chapter, we draw on some of these statistical sources to provide an overview of recent developments in the Nordic countries, also in comparison with the rest of Europe and the OECD area. While offering comprehensive information, these sources also have their shortcomings. First, the provided information is basically static in nature in the sense that it gives a snapshot of the situation in – typically – a specific year, and compares the findings with corresponding information from a previous year (given that such information is available). According‐ ly, this type of year‐specific information highlights the average situation at


21

a given point in time, but can say nothing about the underlying dynamics such as the shifts over time of young people between education, work, unemployment and inactivity. Second, indicators aiming to describe at least some specific dynamic aspect of this continuous transition process can usually be derived only for a limited number of countries with the Nordic countries mostly being surprisingly weakly represented. This is the prevailing situation also in specific studies, of which Quin‐ tini and Manfredi (2009) is an illustrative example. They address a topi‐ cal issue – the dynamic nature of youth labour market situations and key pathways of youth leaving secondary education – but, due to data limita‐ tions, they can only include Denmark and Finland out of the Nordic countries and only up to the year 2001. Simultaneously, this highlights the lack of extensive cross‐country comparative evidence on the dynam‐ ic nature of the labour market situation of youth in the Nordic countries and the key pathways followed by school‐leaving youth having acquired different levels of education – not merely a secondary diploma as in the Quintini and Manfredi (2009) study. Indeed, this is the kind of analysis that we will undertake in the subsequent chapters of this report. A key challenge when addressing the issue of young people’s unem‐ ployment problems and risks of marginalisation is that the youth popu‐ lation is far from homogenous. School‐leavers are equipped with differ‐ ent quantities and qualities of formal education. They also differ in a multitude of other dimensions, including school experiences and early lifetime experiences, notably family background. Their transition from school to work and initial labour market experiences reveal considera‐ ble variation both in length and quality. This variation has been shown to partly originate in differences in the school‐leavers’ educational and social background (e.g. Markussen, 2010). However, it has also been maintained to largely reflect the functioning of the labour market, that is, the labour market institutions in force and the labour market policies pursued in a rapidly changing economic and social environment. Indeed, several reports have shown that the conditions for young people to es‐ tablish themselves in the labour market reveal important differences also across the Nordic countries (see e.g. Olofsson and Wadensjö, 2007; Olofsson and Panican, 2008). Furthermore, while early unemployment is known to affect the youth to a substantial degree, it is also recognized that the short‐ and long‐ term consequences of early unemployment differ markedly across young individuals. As pointed out in the outset, a growing body of litera‐ ture – with evidence emerging also for the Nordic countries – indicates that spells of unemployment entail the risk of creating permanent scars

22


especially for disadvantaged youth who tend to be particularly ill‐ prepared for today’s labour market. Our subsequent comparisons, based on available international data, clearly illustrate how important it is to distinguish between young peo‐ ple’s different situations and composites of activities also when undertak‐ ing international comparisons. In particular, since young people often combine school with part‐time work, especially in the Nordic countries, the results obtained vary substantially with the way in which the status of youth is measured. Let us start by first reflecting on youth employment.

1.3.1

Youth employment

Our national longitudinal datasets cover three full youth cohorts: all young people turning 16 in 1993, 1998 or 2003 (see Chapter 2). We therefore start by illustrating – using OECD statistics – the overall em‐ ployment situation for young people five years later, that is, when our young people turned 21. By this age, most of them could be expected to have completed an upper secondary degree. What actually happened to the young people contained in our three youth cohorts will become evi‐ dent in the next chapter. Table 1.1 presents OECD employment population ratios for the Nor‐ dic countries for four selected years: in 1998, 2003 and 2008, as well as in the last year of available data, 2013. In all these years, Denmark shows up with the highest employment population ratio, both for the very young and for the 20–24 year‐olds, followed by Norway, Finland and Sweden. However, the table also reveals that Denmark experienced a marked decline in employment population ratios up to 2003, that Fin‐ land saw a clear improvement in employment population ratios between 2003 and 2008, and that the employment population ratios of Norwe‐ gian males deteriorated in 2003 but recovered by 2008, whereas the employment population ratios of Swedish males started increasing al‐ ready in 1998. In all four countries, there was a remarkable drop in the employment population ratios of young people between 2008 and 2013. This decline was notably stronger for young men than for young women.


23

Table 1.1: Employment population ratios (%) for Nordic youth in 1998, 2003, 2008 and 2013 1998

2003

2008

2013

15–19 year‐olds Denmark Denmark Finland Finland Norway Norway Sweden Sweden

males females males females males females males females

20–24 year‐olds

Denmark Denmark Finland Finland Norway Norway Sweden Sweden


2008–2013 %‐points

%

57.9 60.2 25.7 19.9 42.2 42.3 21.7 25.1

51.4 49.5 24.7 23.3 38.4 43.0 25.8 32.2

59.2 59.1 26.6 27.5 42.1 46.1 20.3 26.0

41.1 47.1 17.0 26.6 33.8 39.5 17.5 24.1

‐18.1 ‐12.0 ‐9.6 ‐0.9 ‐8.3 ‐6.6 ‐2.8 ‐1.9

‐30.6 ‐20.3 ‐36.1 ‐3.3 ‐19.7 ‐14.3 ‐13.8 ‐7.3

75.6 71.2 60.2 48.7 74.3 65.0 58.2 52.5

71.9 65.4 61.8 54.7 68.8 65.2 63.1 57.3

76.3 71.9 70.1 62.3 74.8 71.5 66.8 59.8

63.4 62.6 54.5 60.5 67.4 67.6 59.4 58.0

‐12.9 ‐9.3 ‐15.6 ‐1.8 ‐7.4 ‐3.9 ‐7.4 ‐1.8

‐16.9 ‐12.9 ‐22.3 ‐2.9 ‐9.9 ‐5.5 ‐11.1 ‐3.0

Note: The employment population ratio measures the employed as a percentage of the population in the age group. Source: OECD Labour force statistics. OECDiLibrary.

Figure 1.1 extends the picture by showing employment population ratios of youth aged 15–24 for the Nordic countries in comparison with select‐ ed large OECD economies (with the information referring to the 1st quarter of 2012). The employment population ratios displayed in the figure are measured in two different ways: the vertical axis gives the employment population ratio for all young people, whereas the horizon‐ tal axis restricts the employment ratio to non‐students only, as youth employment ratios can be criticised for being blurred by pupils and stu‐ dents working part‐time while in school. This setting reveals that Den‐ mark, Iceland and Norway have high youth employment ratios, together with Germany, the UK and USA, also when account is made for part‐time working students. Finland and Sweden, on the other hand, move from above the OECD average to below the OECD average when the youth employment ratio is adjusted with respect to students working on a part‐time basis.

24


Figure 1.1: Employment population ratios (%) for all young people aged 15–24 (vertical axis) and separately for non‐students only (horizontal axis), 2012 (1st quarter)

Note: The employment population ratio measures the employed as a percentage of the population in the age group. Source: OECD based on national Labour Force Surveys.

1.3.2

Youth unemployment

Unemployment is a much more widespread phenomenon among youth than among adults virtually everywhere. Young people are also the first to be hit by rising unemployment, and typically the last to benefit from recovering labour markets. Table 1.2 shows unemployment rates, meas‐ ured as a percentage of the labour force, for the four Nordic countries under study in the years 1998, 2003, 2008 and 2013. Among young peo‐ ple under age 20, Denmark stands out with a comparatively low unem‐ ployment rate in 1998, whereas Finland was still struggling with high post‐recession youth unemployment. The unemployment rate of the very young has been increasing in all four countries from 1998 to 2013, except for Norway where it has declined, but only among young women. The unemployment rates of young people aged 20–24 are also high, but not as dramatically high as for those aged 15–19. Yet, in both Finland and Sweden, about one‐fifth of the male labour force of 20–24 year‐olds was unemployed in 2013, according to OECD statistics.


25

Table 1.2: Unemployment rates (%) for Nordic youth in 1998, 2003, 2008 and 2013 1998

2003

2008

2013


2008–2013 %‐points

%


8.0 9.3 24.6 33.3 13.2 13.7 21.1 20.6

9.9 10.7 25.9 31.5 17.3 14.8 18.6 18.3

7.9 10.9 23.3 25.4 13.2 10.5 32.5 31.8

17.9 13.9 34.1 27.6 13.3 11.3 39.3 34.5

10.0 3.0 10.8 2.2 0.1 0.8 6.8 2.7

126.6 27.5 46.4 8.7 0.8 7.6 20.9 8.5


5.7 6.4 18.0 20.0 7.0 7.1 16.2 14.3

8.6 8.1 16.7 16.3 10.4 8.3 12.6 9.9

7.0 6.6 11.5 11.4 4.9 4.1 13.9 14.0

11.8 10.3 19.2 11.7 9.2 5.6 20.1 17.0

4.8 3.7 7.7 0.3 4.3 1.5 6.2 3.0

68.6 56.1 67.0 2.6 87.8 36.6 44.6 21.4


Notes: The unemployment rate measures the number of unemployed as a percentage of the labour force in the age group. Note also that Sweden undertook, in 2007, a harmonisation of the unem‐ ployment definition towards the one used by ILO. This harmonisation seems to be overlooked in the OECD statistics, implying that the Swedish figures for 1998 and 2003 are not directly comparable with those for 2008 and onwards. Based on harmonised data, the unemployment rate of 15–19 year‐old males and females was in 2003 27.1% and 23.8%, respectively. These rates notably exceed those given above. The discrepancy is much lower for 20–24 year‐olds: an unemployment rate of 14.8% for men and 12.5% for women according to harmonised data published by the Swedish statistical bureau SCB. Source: OECD Labour market statistics. OECDiLibrary.

The unemployment rates reported in Table 1.2 are commonly used in international comparisons of youth unemployment. They are based on Labour Force Survey (LFS) data and measured in a traditional way: by relating the number of unemployed to the total labour force (employed + unemployed). This ILO–LFS measure of unemployment is based on ques‐ tions in the LSF asking the interviewees, inter alia, whether they have been actively looking for a job during the past four weeks. The LFS measure of the total number of unemployed may, as a consequence, also include young people who are actively looking for a job while also being enrolled as full‐time students. Another way of measuring unemployment is to rely on registered un‐ employment, which merely comprises those who have registered as unemployed jobseekers at an unemployment office (PES). The extent to which registered unemployment includes full‐time students depends on the eligibility conditions of young people for registering as unemployed jobseekers. Table 1.3, which is reproduced from Halvorsen et al. (2012), shows the percentage share of youth in each age group registered as

26


unemployed in the years 2000 and 2005–2009. We note that the differ‐ ence between registered unemployment and LFS unemployment is par‐ ticularly large for young people, especially for Sweden. A major reason for this discrepancy is that registered unemployment is typically lower among young people not eligible for unemployment benefits. Table 1.3: Registered unemployment rates (%) for Nordic youth, 2000 and 2005–2009 Age

2000

2005

2006

2007

2008

2009

1.0 4.7 5.3

0.8 3.7 4.2

0.7 2.7 2.9

0.7 2.5 2.1

1.0 4.7 3.6

Denmark 16–19 20–24 16–64 Finland 16–19 20–24

3.2 9.4

2.5 9.0

2.4 8.6

2.4 8.1

2.7 8.4

3.5 10.9

Iceland 16–19 20–24 16–70

0.7 1.1 1.1

0.9 2.4 1.7

0.5 1.4 1.1

0.4 1.0 0.5

0.9 2.1 1.4

3.6 9.9 6.8

Norway 16–19 20–24 16–74

1.2 3.3 2.7

1.2 4.5 3.5

1.0 3.2 2.6

0.7 2.2 1.9

0.7 2.0 1.7

1.0 3.5 2.7

Sweden 16–19 20–24 16–64

1.5 5.0 4.1

2.2 5.7 4.2

1.8 5.9 3.6

1.6 5.0 2.9

1.8 4.0 2.8

2.9 6.0 4.0

Notes: The unemployment rate measures the number of registered unemployed as a percentage of the population in the age group. Information for the year 2000 is missing for Denmark while infor‐ mation on all age groups combined is not available for Finland. Source: Halvorsen et al. (2012).

Also the information on unemployment contained in our national longi‐ tudinal datasets used in the subsequent chapters refers to registered unemployment. However, it differs from the information underlying the registered unemployment reported in Table 1.3, as our datasets are re‐ coded to identify all young people enrolled in education as full‐time stu‐ dents also when they appear as employed or unemployed in the original data (see further Chapter 2). Indeed, as in the case of youth employment, the way full‐time students are treated makes a big difference also when calculating youth unemployment rates. We illustrate this by going back to the LFS data to show what happens to the ranking of the Nordic coun‐ tries depending on the unemployment measure used. We thereby start by showing, in Figure 1.2, the number of students as a percentage share of the population in the age group. In the LFS data, more than 80% of those aged 16–19 report studying to be their main activity. For those aged 20–24, this share is notably lower, ranging from


27

40 to almost 60% with Denmark having the highest and Sweden the lowest share of young people reporting studying as their main activity. In view of these large shares of students in both youth populations, it is obvious that the results obtained depend critically on the way students are treated in the calculations. Figure 1.2: Students as a share (%) of the population in the age group, for four Nordic countries

Note: Students are defined as young people who report studying to be their main activity in the LFS. The drops in the series in 2005 for Norway and Sweden are most likely due to data issues. Cf. the notes in Table 1.2. Source: Own calculations based on LFS quarterly data.

Let us now return to unemployment. Figure 1.3 provides information on youth unemployment from the 1st quarter of 2012 for the Nordic coun‐ tries as well as for a selected number of non‐Nordic countries. The hori‐ zontal axis measures the unemployment rate, that is, the number of un‐ employed as a percentage of the labour force. Sweden comes out with a higher youth unemployment rate than the average for Euro countries (EURO). Sweden, together with Finland, also ranks higher than both the UK and the USA. The vertical axis, in turn, measures youth unemployment by means of the unemployment ratio, that is, the number of unemployed as a per‐ centage of the whole youth population. While 23% of the European youth labour force is unemployed, these unemployed young people con‐ stitute less than 10% of the youth population. In terms of the unem‐

28


ployment ratio, however, all Nordic countries except Norway score higher than the average for Euro countries. In both Iceland and Sweden, youth unemployment is, in fact, higher than in Italy when measured by means of the unemployment ratio. Figure 1.3: Youth unemployment rates (%) and unemployment ratios (%), the Nordic countries and selected non‐Nordic countries, 2012 (1st quarter)

Notes: Unemployment rate = unemployed 15–24 year‐olds in relation to the labour force of the age group. Unemployment ratio = unemployed 15–24 year‐olds in relation to the whole population of the age group. Source: OECD as reproduced from Albæk et al. (2014).

However, many of the young people recorded as unemployed in the LFS are actually attending school and, moreover, typically on a full‐time ba‐ sis. Figure 1.4 shows what happens when we remove from the pool of young unemployed those who report studying as their main activity. The vertical axis of Figure 1.4 now shows unemployed young persons who are not attending school. For the Euro area as a whole, the youth unem‐ ployment ratio drops from 9.3% to 7.2% of the youth population. For the Nordic countries, the change is even larger. After this correction, all Nordic countries rank among those with the lowest level of youth un‐ employment (among the non‐students). The countries with the largest drop below the 45‐degree line added to Figure 1.4 are the ones with the largest proportion of unemployed youth who are also students. The pattern displayed in the figure thus arises from the fact that the unemployed young persons who are not at the same time attending school make up a smaller proportion in the Nordic countries


29

compared to the other countries in the figure. In all Nordic countries, the proportion of unemployed young people who are not also in school is less than one‐half, and as small as one‐third in Sweden. The huge difference between youth unemployment ratios including and excluding unemployed students is also evident in the next two fig‐ ures which, moreover, separate 16–19 year‐olds from 20–24 year‐olds. A comparison of Figure 1.5a with Figure 1.5b reveals a strikingly large drop in the unemployment ratio for the very young (16–19 year‐olds) when leaving out students recorded as being unemployed. Figure 1.4: Youth unemployment ratios (%) with and without students, the Nordic countries and selected non‐Nordic countries, 2012 (1st quarter)

Notes: Unemployment ratio = unemployed 15–24 year‐olds in relation to the whole population of the age group. Unemployment ratio, non‐students = unemployed 15–24 year‐olds with studying not being their main activity, in relation to the whole population of the age group. The red line shows the 45 degree angle. Source: OECD, as reproduced from Albæk et al. (2014).

30


Figure 1.5a: Youth unemployment ratios (%) for four Nordic countries, 1995–2012

Notes: Unemployment ratio = unemployed in relation to the whole population of the age group. Cf. the notes in Table 1.2 concerning the unemployment information on Sweden prior to 2005. Source: Own calculations based on LFS quarterly data.

Figure 1.5b: Youth unemployment ratios (%), excluding students, for four Nordic countries, 1995–2012

Notes: Unemployment ratio = unemployed in relation to the whole population of the age group. Students are defined by self‐reported main activity. Cf. the notes in Table 1.2 concerning the unem‐ ployment information on Sweden prior to 2005. Source: Own calculations based on LFS quarterly data.


31

1.3.3

NEET rates

NEETs refer to young people who are Not in Employment, Education or Training. Figure 1.6 gives NEET rates for the last quarter of 2012 for 15–24 year‐olds in OECD countries, measured as a percentage of the total population in the age group. The Nordic countries are located rela‐ tively high up on this ranking scale, implying that they are characterised by low NEET rates. For 2012, the share of young people in NEET activi‐ ties was estimated to be 5.9% for Denmark, 6.7% for Norway, 7.2% for Sweden and 8.4% for Finland. All in all, the information provided so far shows that the labour mar‐ ket is quite accessible for young people in the Nordic countries. In par‐ ticular, employment rates are relatively high among those who do not attend school and, conversely, unemployment rates are relatively low. The descriptive evidence presented above also implies that the attach‐ ment to the labour market is relatively strong also among young people attending school. This follows from the fact that large shares of young people recorded to be employed or unemployed are, actually, pupils and students with studying as their main activity. Simultaneously, they are an integral part of the labour force. Figure 1.6: NEET rates (%) for 15–24 year‐olds in OECD countries, as split by unemployment and inactivity, 2012 (4th quarter)

Notes: NEET rate = young people not in employment, education or training as a percentage of the population in the age group. Source: OECD Society at a Glance 2014.

32


1.3.4

Upper secondary education – completion, non‐ completion and dropout

Finally, we take a closer look at school completion, non‐completion and dropout with the focus being on upper secondary education. The pro‐ portion of 30–34 year‐olds holding an upper secondary degree is not particularly high in Denmark or Norway, as is evident in Figure 1.7. Fin‐ land and Sweden, on the other hand, rank quite high with more than 90% of the 30–34 year‐olds holding an upper secondary degree in 2012. If we explore the average age of upper secondary graduation (Figure 1.8), Sweden comes out with the lowest average age among the Nordic countries, or about 18 years‐of‐age. In the other Nordic countries, the average graduation age is much higher for both general programs and especially for vocational programs. Indeed, the average age of upper secondary graduation from vocational programs is 28 in both Denmark and Norway. For Finland, it is reported to be even higher. Figure 1.7: Attainment (%) of an upper secondary degree among 30–34 year‐ olds in OECD countries, 2012

Source: OECD Education at a Glance 2014.


33

Figure 1.8: Average age of upper secondary graduation in OECD countries, 2012

Source: OECD Education at a Glance 2014, Chart A2.2.

There are numerous ways to measure dropout or non‐completion rates from upper secondary education. Figure 1.9 illustrates the share of stu‐ dents who successfully completes after the theoretical duration of the upper secondary program, or two years after the theoretical duration of the program. The countries are sorted according to the successful com‐ pletion of girls in upper secondary programs. All Nordic countries rank below the OECD average, with Sweden and Finland doing slightly better than Denmark and Norway.

34


Figure 1.9: Upper secondary completion rates (%) in OECD countries, by gender

Source: Education at a Glance 2014. Chart A2.5.

Table 1.4, finally, presents alternative measures of completion rates, as reported by Eurostat and OECD. While Figure 1.9 gives the completion rates among those who attend upper secondary programs, Table 1.4 presents completion rates as percentage shares related to the whole youth population. Table 1.4: Completion, early leaving and dropout rates (%) for the Nordic countries Denmark

Finland

Iceland

Norway

Sweden

85.0 11.0 96.0 9.1 8.9

70.4 17.4 87.8 23.2 20.1

77.7 11.8 89.5 18.4 14.8

75.4 0 75.4 8.0 7.5

9.7

55.2

20.3

7.4

Upper secondary graduation rates 3 years

20% 10% 0% Denmark

Finland

Norway

Sweden

Note: The figure is based on the information provided in Tables 5.2 to 5.6.


167

The information provided in Figure 5.6 is presented in a rescaled format in Table 5.7 in order to illustrate the corresponding situation when ac‐ count is made for the fact that the non‐completion rate among 21‐year‐ olds differs substantially across the four countries (see Chapter 2). In both Denmark and Norway, more than one‐fifth of the youth population con‐ tinue in post‐compulsory education for at least 3 years but fail, nonethe‐ less, to achieve an upper secondary degree by age 21. The corresponding share for Finland and Sweden is just above 10%. The cross‐country situa‐ tion is much more similar when, instead, comparing the share of young‐ sters dropping out early from post‐compulsory education: almost 14% of Danish youth follow early school‐leaving tracks, compared to about 8% for Finland and Norway, and close to 6% for Sweden. Table 5.7: Distribution of non‐completers across stylized pathways by number of initial years spent in post‐compulsory education before dropping out, for the four Nordic countries, %‐share of the full youth population Number of initial years in post‐ compulsory education

> 3 years 3 years 2 years 1 year Non‐starter Non‐completers’ share in the full youth population

%‐share of the full youth population Denmark

Finland

Norway

Sweden

14.6 7.0 7.1 5.9 2.8 37.3

6.3 4.1 2.7 2.4 2.5 18.0

10.8 10.1 5.6 2.4 0.8 29.7

5.9 5.2 2.3 1.2 1.4 16.0

Notes: The percentage shares displayed in Figure 5.6 as recalculated in relation to the full youth population of each country.

Next, we turn from the activity dominating the start of each stylized pathway to the activity dominating the end years of each pathway. As shown in Figure 5.7, the stylized post‐compulsory‐school pathways dominated by continuous engagement in full‐time education cover more than one‐third of the non‐completers in all four countries. These young people spend most of their years from age 16 up to age 20 as full‐time students but fail, nonetheless, to finalise their upper secondary educa‐ tion by age 21. The share of young non‐completers spending most of their early post‐compulsory‐school years in education is largest in Den‐ mark (39%) and lowest in Finland (32%). Denmark also has the largest share of young non‐completers following post‐compulsory‐school tracks moving them early into working life (close to 36%), followed by Finland (30.5%). In Norway and Sweden, young non‐ completers are less likely to leave school for employment. Taken together, these early education‐ and employment‐dominated pathways comprise 75% of the Danish non‐completers. The corresponding share of non‐

168


completers spending most of their years from age 16 up to age 20 in either school or work is 66% for Finland, 64% for Norway and 62% for Sweden. Figure 5.7: Distribution (%‐share) of young non‐completers across stylized pathways by main activity after leaving post‐compulsory education, for the four Nordic countries

100% 90% 80% 70% 60% 50% 40% 30% 20% 10%

Other Disability benefits Unemployment Employment Education

0% Denmark

Finland

Norway

Sweden

Note: The figure is based on the information provided in Tables 5.2 to 5.6.

Swedish non‐completers are more likely to spend time in (registered) unemployment after having dropped out from upper secondary educa‐ tion than are non‐completers in the other three Nordic countries: an average of close to 17% of Swedish non‐completers end up in unem‐ ployment tracks before turning 21, compared to about 12% for Finland, some 9% for Norway and less than 5% for Denmark. Most Swedish non‐ completers in unemployment tracks have spent three years in upper secondary education before showing up as unemployed jobseekers. The relatively large share of Finnish non‐completers in unemployment tracks is mainly due to the weak employment prospects of low‐educated youth who tried to enter the labour market in the high‐unemployment years of the 1990s. In the two younger Finnish cohorts, the shares of non‐completers experiencing early spells in unemployment are similar to those of Danish and Norwegian non‐completers. Relatively few non‐completers follow pathways shifting them already at an early age into disability arrangements. The highest share (an aver‐ age of 4%) is observed for Sweden. Much larger shares of young non‐ completers show up in trajectories dominated by (unknown) activities outside both education and the labour force. This share is notably high


169

(close to 25%) for Norway. In the other three countries, it is clearly low‐ er but still of a remarkable size: about 19% for Denmark and Finland, and some 17% for Sweden. Consider next the same distribution, now related to the full youth pop‐ ulation of each country. After this rescaling, we find that about 13% of Danish youth shift early into employment tracks, which typically means that they lack an upper secondary degree still when aged 21. In Norway, the corresponding share is 8%, in Finland 6% and in Sweden only 4%. Denmark and Norway also come out with a larger share of the youth pop‐ ulation (7.3%) following early tracks ending with the young person with‐ drawing from both education and the labour market already before turn‐ ing 21. In Finland and Sweden, such tracks are much less common. When related to the full youth population, the country‐specific shares of youth ending up, already at an early age, in unemployment or disability become quite similar in size. Indeed, Finland, Norway and Sweden have about the same share of youth (2.4–2.8%) moving early into registered unemploy‐ ment – and non‐completion of an upper secondary degree. The share for Danish youth is smaller, 1.7%, but not as different as the comparison across non‐completers suggests (in Figure 5.7). Table 5.8: Distribution of young non‐completers across stylized pathways by main activity after leaving post‐compulsory education, for the four Nordic countries, %‐share of the full youth population Main activity after leaving post‐compulsory education Continue in education Employment Unemployment Disability benefits Other (inactivity) Non‐completers’ share in the full youth population

%‐share of the full youth population Denmark

Finland

Norway

Sweden

14.6 13.3 1.7 0.5 7.3 37.3

5.8 6.3 2.4 0.5 3.0 18.0

10.8 8.2 2.8 0.6 7.3 29.7

5.9 4.0 2.7 0.6 2.8 16.0

Notes: The percentage shares displayed in Figure 5.7 as recalculated in relation to the full youth population of each country.

Hence, the main implication of the large difference among the 21‐year‐ olds in non‐completion rates between especially Denmark, but also Norway, on the one hand, and Finland and Sweden, on the other hand, is that a notably larger share of Danish and Norwegian youth continue straight in post‐compulsory education for typically three or more years, but without completing an upper secondary degree by age 21, or leave school for work before having graduated. These differences are likely to be basically due to the institutional setting, in particular the organisation of vocational training. The shares of early dropouts going into typical NEET paths are more similar across the four countries.

170


6. Labour market outcomes as young adults Our analyses have so far focused on exploring young people’s main ac‐ tivities and school‐to‐work‐transition patterns over the five years fol‐ lowing upon completion of compulsory school, that is, from age 16 up to age 20. A logical next step is to ask: What happens to these youngsters after they have turned 20? What kind of main activities – studying, em‐ ployment, unemployment, disability arrangements or other types of inactivity – are they mostly engaged in as young adults? Can we observe distinct and rather stable differences in this respect across genders and/or between those differing in their educational background? Or is it possible that these later outcomes are, by and large, quite similar for young men and women, as well as for early and later completers of a post‐compulsory educational degree and, possibly, even for non‐ completers, i.e. those with no exam beyond primary education still as a young adult? Last, but not least, can we identify clear‐cut similarities or dissimilarities in all these important dimensions across the four Nordic countries under study? This chapter sets out to provide answers to these key questions.

6.1 Main activities beyond age 21 – all young people We start by recalling the overall pattern of labour market outcomes for young adults based on pooled information on our three country‐specific youth cohorts (16‐year‐olds in 1993, 1998 and 2003). In other words, we first re‐report in which main activities the young people covered by our national datasets are engaged at three different ages – 21, 26 and 31. But instead of repeating the country‐specific graphs contained in Figures 2.1a to 2.1d of Chapter 2, we now present the same information from a slightly different angle. Then we refine this descriptive information in an attempt to unravel to what extent the economic situation is possibly reflected in the labour market outcomes of our youth cohorts under scrutiny.

6.1.1

Reproducing the general picture from a different angle

In particular, Figure 6.1 contains three graphs with the first graph providing information on the allocation of each country’s young people across our five main activity categories – studying, employment, unem‐ ployment, disability arrangements and other types of inactivity – when they turned 21. The next (middle) graph gives the corresponding infor‐ mation five years later, at age 26. The bottom graph, finally, displays the situation ten years later, at age 31. Each graph then highlights cross‐ Nordic similarities and dissimilarities in young people’s activity shares at these particular ages. A top‐down comparison of the three age‐specific distributions sheds, in turn, light on the changes in activity shares within countries when moving from age 21 to age 31. Indeed, this line‐of‐ comparison is identical to the information provided in the country‐ specific Figures 2.1a to 2.1d of Chapter 2. Figure 6.1: Distribution (%‐share) of young people across main activities at age 21, 26 and 31, respectively, based on pooled information on all three youth co‐ horts, by country

Age 21 100% 80% 60% 40% 20% 0% Denmark Student

172

Finland

Employed

Norway

Unemployed


Pensioner

Sweden Other


Finland

Employed

Norway

Unemployed

Pensioner

Sweden Other


Finland

Employed

Norway

Unemployed

Pensioner

Sweden Other

Note: For definitions of the five main activity groups, see Chapter 2.

Since Figure 6.1 contains the same basic information at ages 21, 26 and 31 as Figures 2.1a to 2.1d, albeit in a different mode, we here comment only briefly on the patterns displayed in the figure. First, the overall pic‐ ture looks much the same across the four Nordic countries under study: rapidly declining shares in full‐time education and growing shares in employment when young people grow older. This is, in effect, the most conspicuous change in activity shares occurring beyond age 21. Second, in all four countries there are non‐negligible shares of young people experiencing unemployment, health problems moving them onto disability benefits, or other forms of inactivity excluding them from both


173

education and working life. Although also these NEET shares change over time, the changes are rather small in magnitude with no systematic pattern discernible across the four countries. Third, the situation at age 21 looks more or less the same in all four countries when adding up young people’s shares in full‐time education and employment, on the one hand, and in NEET activities, on the other hand. Moreover, this overall pattern is only marginally different five years later, at age 26. While the distribution of young people across main activities appears to reveal more distinct cross‐country differences by age 31, when compared to the situation five or ten years earlier, we need to recall that the outcome at age 31 is based on information on one sin‐ gle cohort, viz. the oldest cohort representing young people who turned 16 in 1993.

6.1.2

Cohort effects of the economic situation

Next we refine the descriptive information provided in Figure 6.1. This exercise departs from the fact that we have information on later labour market outcomes for two of our three youth cohorts under scrutiny: the 1998 cohort of 16‐year‐olds can be followed up to age 26 (in 2008) while the 1993 cohort of 16‐year‐olds can be traced up to age 31 (in 2008). A main motivation for looking somewhat closer into these two cohorts is that they started their school‐to‐work transition in very dif‐ ferent economic contexts. As referred to in the outset, we know from the literature that the prevailing economic situation tends to have far‐ reaching career consequences for young labour market entrants. Ac‐ cordingly, we might expect the labour market experiences in adulthood to be different for these two cohorts. In Figure 6.1, the distribution across activities of our young people when aged 26 shows the average outcome for the 1993 and 1998 co‐ horts. In contrast, and as already underlined above, the distribution of 31‐year‐olds reflects the situation for the oldest (1993) cohort only. By splitting the information provided in the figure for age 26, we may un‐ dertake two additional comparisons potentially shedding light on the following questions: Does the distribution across main activities at age 26 look different for the “economic‐bust” cohort of 1993 and the “eco‐ nomic‐boom” cohort of 1998? Does the situation change markedly be‐ tween age 26 and age 31 for the economic‐bust cohort of 1993, or does it remain approximately unchanged? These two comparisons are under‐ taken in Figures 6.2 and 6.3. Since the economic recession in the early

174


1990s hit all four Nordic countries under study, we would expect cross‐ cohort differences to show up for all four countries. According to Figure 6.2, the share of young people engaged in either full‐time education or employment is, by age 26, slightly higher in the 1998 “economic‐boom” cohort than in the 1993 cohort. However, this holds true for Denmark and Finland only. In Denmark, this higher “activ‐ ity” share of the 1998 cohort is due to full‐time studying still at age 26 being more common in the 1998 cohort, whereas the share in working life is of the same magnitude in the two cohorts. In Finland, on the other hand, the 1998 cohort is by age 26 more engaged in both full‐time stud‐ ies and employment than was the 1993 cohort when aged 26. The situation looks different for Norway and Sweden. For both coun‐ tries, we observe a small decline across the two cohorts in the share of full‐time students and a slight increase in the share in employment. For Sweden, these two opposite‐signed changes are of much the same mag‐ nitude, for which reason we see principally no difference across the two cohorts in the total share of 26‐year‐olds in either education or em‐ ployment. For Norway, in contrast, the drop in the share of full‐time students more than outweighs the concomitant increase in the share of employed. This results in an “activity” share among 26‐year‐old Norwe‐ gians that is lower in the 1998 cohort than in the 1993 cohort. Figure 6.2 further indicates that the share of 26‐year‐olds in (regis‐ tered) unemployment was clearly lower in the 1998 cohort. This holds true for all four countries and is, most likely, due to a favourable eco‐ nomic situation in combination with increased volumes of active labour market policies. Simultaneously, however, the share of 26‐year‐olds on disability benefits or in other types of inactivity appears to be of much the same size in the two cohorts (Finland) or even larger in the 1998 economic‐boom cohort (notably in Norway, but also in Denmark and Sweden). All in all, Figure 6.2 seems to suggest that the prevailing busi‐ ness cycle has, at most, been reflected in young people’s activities when it comes to studying and working, including unemployment, whereas withdrawal from both education and the labour market is mainly the result of other processes and mechanisms.


175

Figure 6.2: Distribution (%‐share) of young people across main activities at age 26, by country: comparison of the “economic‐bust” cohort of 1993 with the “eco‐ nomic‐boom” cohort of 1998

Denmark 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

1993 cohort Student

Employed

1998 cohort Unemployed

Pensioner

Other

Finland 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1993 cohort Student

176

Employed



Pensioner

Other

Norway 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Student 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Student

1993 cohort Employed


Pensioner

Other

Sweden

1993 cohort Employed


Pensioner

Other


Figure 6.3, finally, shows that the overall labour market outcome of the 1993 “economic‐bust” cohort improved over the 5‐year period from age 26 up to age 31. In Sweden, more than eight out of ten (82%) of the co‐ hort’s young people were working at age 31 with an additional 9% still being enrolled in education. Hence, 91% of the Swedes belonging to this cohort were either studying or working when aged 31 (compared to close to 89% when aged 26). In other words, the situation of the Swedish 1993 cohort had, by age 31, turned very similar to the situation of the Danish 1993 cohort in terms of both employment and education. Also


177

the Finnish 1993 cohort had by age 31 experienced a notable improve‐ ment in its employment situation which, nonetheless, remained notably weaker (73%) than for the Danish and Swedish 1993 cohorts. However, this lower employment level is not necessarily entirely due the economic recession and the high unemployment levels of the 1990s: as has be‐ come evident also in the previous chapters, Finland is throughout char‐ acterised by a lower share of employed, when compared to the other three Nordic countries under study. Figure 6.3: Distribution (%‐share) of young people across main activities at age 26 and 31, respectively, by country: the “economic‐bust” cohort of 1993

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Student 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Student

178

Denmark

Age 26 Employed

Age 31 Unemployed

Pensioner

Other

Finland

Age 26 Employed

Age 31 Unemployed


Pensioner

Other

Norway 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Age 26 Student

Employed

Age 31 Unemployed

Pensioner

Other

Sweden 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Age 26 Student

Employed

Age 31 Unemployed

Pensioner

Other


In parallel with improving employment up to age 31, the 1993 cohort of Denmark, Finland and Sweden experienced a decline in the share of un‐ employed jobseekers. But simultaneously all three countries also saw an increasing share of the cohort’s young people moving outside both edu‐ cation and the labour market, when comparing the situation at age 26 to that prevailing five years later, at age 31. This increase is, for the most part, explained by a higher share receiving disability benefits. A slightly different pattern emerges for the Norwegian 1993 cohort. While the share of the cohort’s young people moving into working life


179

increased between age 26 and 31 also in Norway, this growth could not compensate for the concomitant drop in the share still enrolled in full‐ time education. As a consequence, the cohort’s “activity” (education + employment) share was marginally lower at age 31 than at age 26. Sim‐ ultaneously, the share of the cohort’s young people moving outside both education and the labour force increased quite markedly, from about 9% at age 26 to almost 14% at age 31. Moreover, only a minor part of this increase is explained by a growing number of the cohort’s young adults moving into disability arrangements. Instead, the main explanation seems to be that they withdraw into other types of inactivity not covered by the large administrative registers from which our national datasets are compiled.

6.1.3

Main findings

In this sub‐chapter, we have addressed two interconnected questions: Does the distribution across main activities at age 26 look different for the “economic‐bust” cohort of 1993 and the “economic‐boom” cohort of 1998? Does the labour market situation change between age 26 and age 31 for the economic‐bust cohort of 1993, or does it remain approximate‐ ly unchanged? Our descriptive analysis based on information for the full youth co‐ horts cannot provide clear‐cut answers to these questions in the sense that we see no systematic cross‐country trends in activity shares at age 26 for the economic‐bust and economic‐boom cohorts. In particular, the share in employment by age 26 is slightly higher in the 1998 cohort than in the 1993 cohort for Finland, Norway and Sweden, but not for Den‐ mark. The share enrolled in full‐time education still at age 26 is higher in the Danish and Finnish 1998 cohorts, whereas the opposite holds true for Norway and Sweden. Adding up the shares in education and em‐ ployment reveals no conspicuous differences between the two cohorts, either. Instead, we see a slight increase in this “activity” share across the Danish and Finnish cohorts, no change across the Swedish cohorts and a decline across the Norwegian cohorts. While the share in (registered) unemployment at age 26 is, indeed, lower in the 1998 cohort in all four countries, this change seems to have occurred at the expense of a much higher (Norway) or slightly higher (Denmark and Sweden) share of 26‐ year‐olds in the 1998 cohort standing outside both education and the labour force. For Norway, the results further indicate that this conspicu‐ ous increase in the share of young people withdrawing from both educa‐

180


tion and the labour force has been fed not only by a flow from unem‐ ployment but also from full‐time education. The country‐specific patterns are more similar when comparing the labour market situation of the 1993 cohort of 16‐year‐olds at two points later in life, viz. at age 26 and age 31. In all four countries, we observe an improvement in the cohort’s labour market situation between age 26 and 31 in terms of more employment and less unemployment. Finally, while the share outside both education and the labour force reveals an upward trend in all four countries, this change up to age 31 has general‐ ly been quite modest, except for the Norwegian 1993 cohort. However, these weak signs of the deep economic recession of the ear‐ ly 1990s having had an impact on the labour market outcomes of the 1993 cohort of 16‐year‐olds do not mean that their situation was unaf‐ fected by the difficult employment situation that prevailed for several years also after the start of the economic recovery. First and foremost, we need to recall that our data measures registered unemployment. As shown and discussed at length in Chapter 1, this measure is likely to underestimate the prevalence of unemployment among young people: many of them do not fulfil the conditions for signing on, or then they may choose not to register if not being eligible for receiving unemploy‐ ment benefits. However, it might also be that the economic crisis affect‐ ed specific groups of young persons, instead of influencing all young people about to enter the labour market. In Finland, for instance, the recession started in male‐dominated export industries, which resulted in surging male unemployment rates and widespread destruction of espe‐ cially low‐skilled jobs often occupied by low‐educated youngsters. Next, we therefore deepen our analysis in two respects: first, by comparing the situation of young men and women, and second, by contrasting the outcomes of completers and non‐completers of an upper secondary de‐ gree by age 21.

6.2 Main activities beyond age 21 – young men vs. young women Next, we compare the labour market outcomes in adulthood across gen‐ ders: Are the labour market experiences of young men and young wom‐ en distinctly different at age 21, 26 and/or 31? Can we identify a strong‐ er impact on either gender of the economic crisis in the early 1990s?


181

6.2.1

Labour market outcomes by gender at three specific age points

A logical way to start this male–female comparison is to split the infor‐ mation provided in Figure 6.1 by gender. This is done in Figure 6.4 with the structure of the figure now emphasising, first and foremost, within‐ country differences rather than between‐country differences. Figure 6.4: Distribution (%‐share) of young people across main activities at age 21, 26 and 31, respectively, based on pooled information on all three youth co‐ horts, by gender and country

Denmark 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Men at 21 Student

Women at 21 Employed

Men at 26

Women at 26

Unemployed

Men at 31 Pensioner

Women at 31 Other

Finland 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Men at 21 Student

182


Men at 26

Women at 26

Unemployed


Men at 31 Pensioner

Women at 31 Other

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Men at 21 Student 100% 80% 60% 40% 20% 0% Men at 21 Student

Norway


Men at 26

Women at 26

Unemployed

Men at 31 Pensioner

Women at 31 Other

Sweden


Men at 26

Women at 26

Unemployed

Men at 31 Pensioner

Women at 31 Other


The first graph of Figure 6.4 highlights the situation for Denmark. We see basically no distinct differences across genders among young Danes, irrespective of whether we compare their situation at age 21, age 26 or age 31. The only divergence in their distributions across main activities concerns the higher share of young women continuing in full‐time edu‐ cation and a correspondingly higher share among young men moving into working life, a difference pointed out already in previous chapters. However, also this cross‐gender difference seems to have almost disap‐ peared by age 31.


183

In the other three countries, there are clearly more differences in the distribution across activities between young men and young women. While much larger shares of young women are in full‐time education at age 21, this difference across genders diminishes steadily over time, but it does not come close to vanishing as in Denmark. Likewise, also the gender gap in employment shares observed at age 21 shrinks over time, but still by age 31 working is more common among men than among women. Nonetheless, taken together, larger shares of young women than of young men are in either education or work. This holds true at all three age points for Norway and Sweden, but only at age 21 for Finland. At age 26, as well as at age 31, the share of Finnish women either working or studying is lower than for Finnish men. These gender differences in education and employment shares are, of course, mirrored by corresponding differences across genders in relation to NEET activities. However, in this respect it is much harder to find com‐ mon patterns for Finland, Norway and Sweden. In particular, while all three countries are characterised by a lower share in unemployment among young women than among young men aged 21, this gap prevails up to age 31 in Finland, disappears by age 31 in Norway, and is reversed by age 31 in Sweden. A similar pattern is, in effect, discernible across gen‐ ders, ages and countries for those on disability benefits. For other types of inactivity, we observe for Finland a larger share for young women than for young men with, moreover, the inactivity share among young Finnish women increasing rapidly with age. The inactivity share is large and grow‐ ing with age also in Norway but, in contrast to Finland, this seems to hold true for both genders. For Sweden, on the other hand, the inactivity share shrinks with age for both genders, more for young women than for young men. By age 31, a clearly larger share of Swedish men than of Swedish women belongs to the dumping category of “other” inactivity.

6.2.2

Gender‐specific cohort effects of the economic situation

As a second step in our cross‐gender comparisons, we repeat the split by cohort of the distributive information provided for age 26 in an attempt to unravel whether or not the economic situation at the time of labour market entry has eventually affected young men and young women dif‐ ferently when it comes to their distributions over main activity catego‐ ries. The results of this exercise are displayed in Figure 6.5.

184


Figure 6.5: Distribution (%‐share) of young people across main activities at age 26, by gender and country; comparison of the “economic‐bust” cohort of 1993 with the “economic‐boom” cohort of 1998

Denmark 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1993 cohort, 1998 cohort, 1993 cohort, 1998 cohort, men men women women Student Employed Unemployed Pensioner Other Finland 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1993 cohort, 1998 cohort, 1993 cohort, 1998 cohort, men men women women Student Employed Unemployed Pensioner Other


185

Norway

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1993 cohort, 1998 cohort, 1993 cohort, 1998 cohort, men men women women Student Employed Unemployed Pensioner Other Sweden 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1993 cohort, 1998 cohort, 1993 cohort, 1998 cohort, men men women women Student Employed Unemployed Pensioner Other Note: For definitions of the five main activity groups, see Chapter 2.

In the previous sub‐chapter focusing on all young people, we noted that the share of young people in either education or employment is some‐ what higher at age 26 in the 1998 “economic‐boom” cohort, but only in Denmark and Finland. For Denmark, this higher “activity” (education + employment) share was found to be due to a larger share of full‐time students in the 1998 cohort, whereas the share in working life appeared to be of the same magnitude in the two cohorts. As is evident in Figure 6.5, this overall pattern shows up for both genders. For Finland, we found a slightly higher share in both education and employment among 26‐year‐olds belonging to the 1998 cohort. Also these patterns are re‐

186


peated for both genders. For Sweden, we identified opposite‐signed but balanced changes in educational and employment shares across the two cohorts, resulting in approximately similar “activity” shares for the 1993 and 1998 cohorts. This pattern is discernible for both genders. For Nor‐ way, finally, we saw a similar but more unbalanced trend across the two cohorts. In particular, the increase in the share of employed was not enough to compensate for the concomitant decline in the share of 26‐ year‐olds enrolled in full‐time studies, which showed up as a lower “ac‐ tivity” share in the 1998 cohort than in the 1993 cohort. Again, the same pattern holds true for both genders. Common to all four countries was a share of 26‐year‐olds in (regis‐ tered) unemployment that was lower in the 1998 cohort. Also in this re‐ spect we observe the same pattern for young men and young women. The cohort‐specific shares of 26‐year‐olds on disability benefits or in other types of inactivity were found to reveal much more cross‐country varia‐ tion: for Finland, the inactivity share in the two cohorts was noted to be of much the same size, but for Denmark, Sweden and notably for Norway, the share of young people outside both education and the labour market was found to be larger in the 1998 than in the 1993 cohort. Again, the same cross‐cohort pattern is discernible for both genders. Hence, the dif‐ ferences between the “economic‐boom” and the “economic‐bust” cohort in the distribution of young people across main activities pointed out in the previous sub‐chapter do not show up differently among young men and young women; the same pattern emerges irrespective of gender. Moreover, this holds true for all four countries. Finally, we take a gendered perspective on the distribution across main activities of the 1993 “economic‐bust” cohort at two age points: when aged 26 and 31, respectively. Here, the main question is whether or not the labour market outcomes of males belonging to this particular cohort possibly evolved differently over these years, when compared to the experiences of their female counterparts. For this purpose, we split the information given in Figure 6.3 above to indicate the situation of young men vs. young women belonging to the 1993 cohort of 16‐year‐ olds. This comparison is undertaken in Figure 6.6. For Denmark, Finland and Sweden we found, based on Figure 6.3, that the overall labour market outcome of this “economic‐bust” cohort improved over the 5‐year period from age 26 up to age 31. In all three countries, about nine out of ten of the cohort’s young people were either studying or working at age 31. Moreover, in all three countries, the share of young people registered as unemployed jobseekers declined up to age 31, whereas the share outside both education and the labour force was


187

marginally higher. Again, much the same overall pattern is discernible for both men and women: a decline in the share enrolled in full‐time studies accompanied by a strong increase in the share in employment, less unemployment but slightly more inactivity. For both genders, this increase in the inactivity share by age 31 seems to be mainly due to a growing inflow into disability arrangements. Figure 6.6: Distribution (%‐share) of young people across main activities at age 26 and 31, respectively, by gender and country, the economic‐bust cohort of 1993

Denmark 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Men at 26 Men at 31 Women at 26 Women at 31 Student Employed Unemployed Pensioner Other Finland 100% 90% 80%

70% 60% 50% 40% 30% 20% 10% 0% Men at 26 Men at 31 Women at 26 Women at 31 Student Employed Unemployed Pensioner Other

188


Norway 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Men at 26 Men at 31 Women at 26 Women at 31 Student Employed Unemployed Pensioner Other Sweden 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Men at 26

Student

Men at 31

Employed

Women at 26

Unemployed

Women at 31

Pensioner

Other


Figure 6.3 mediated a slightly different pattern for the Norwegian 1993 cohort of 16‐year‐olds: despite a growth in the share of employed be‐ tween age 26 and age 31 also in Norway, this increase could not com‐ pensate for the concomitant drop in the share of full‐time students, which resulted in an “activity” share of the 1993 cohort that was lower at age 31 than at age 26. Moreover, while the share in unemployment declined, the share of the cohort’s young people withdrawing from both education and the labour market had increased markedly by age 31. As is evident from Figure 6.6, this overall pattern is strikingly similar for men and women.


189

6.2.3

Main findings

In Chapter 2, we made a first comparison between genders also with respect to their distribution across main activities at age 21. In this sub‐ chapter, we have expanded this comparison to two age points later in life: five years later, at age 26, and ten years later, at age 31. Doing so, we observe basically no differences across genders among young Danes, irrespective of whether we compare their labour market situation at age 21, age 26 or age 31. Indeed, also the higher share of young women con‐ tinuing in post‐compulsory education and the correspondingly higher share among young men moving already at a relatively early age into working life are differences that seem to have disappeared by age 31. We find much more differences in the gender distributions across main activities for the other three countries. The gender gap in the shares of full‐time students and the employed narrows steadily over time but is discernible also at age 31 with enrolment in full‐time educa‐ tion still being more common among young women and working being more common among young men. However, simultaneously we observe larger shares of young men than of young women outside education and work. This holds true at all three age points (21, 26, 31) for Norway and Sweden, but only at age 21 for Finland. At both age 26 and age 31, the share of Finnish women neither working nor studying on a full‐time basis is much higher than for Finnish men. Attempts were also made to identify possible differences in the dis‐ tribution of young men and young women across main activities due to changes in the economic environment. The simple exercises undertaken in this respect do not provide support for changes in the business cycle showing up differently among young men and young women. In other words, to the extent that changing business cycles are reflected in the overall distribution of young people across main activities, men and women seem to be affected in much the same way. This finding emerges for all four countries under study.

6.3 Main activities beyond age 21 – completers vs. non‐completers As a final step in this descriptive analysis of the labour market outcomes of our three youth cohorts when young adults, we turn the focus to a comparison of upper‐secondary‐school completers and non‐completers. Again, completers refer to those young people in our three youth cohorts who succeeded in finalising their upper secondary education within five

190


years after leaving compulsory school. The group of non‐completers, in turn, comprises the remaining young people, that is, those with still no post‐compulsory degree when reaching 21 years‐of‐age. This definition is valid throughout this sub‐chapter. In the next sub‐chapter (6.4), how‐ ever, we will use a less strict definition in the sense that we allow for the fact that at least some of these non‐completers do complete their upper secondary education, but only later on, after the age of 21 (cf. Chapter 2). We start by comparing the distribution of completers and non‐ completers across our five main activity categories when aged 21. This comparison of labour market outcomes is then repeated five years later, at age 26 and, finally, ten years later, at age 31. We conclude by return‐ ing to the question of the role played by the economic situation for later labour market outcomes, with the focus now being on a comparison of completers and non‐completers.

6.3.1

Situation five years later, at age 21

Figure 6.7 illustrates to what extent completers (upper graph) and non‐ completers (lower graph) are engaged in different main activities when aged 21, that is, five years after leaving compulsory school. From the fig‐ ure it is evident that the distribution across main activities varies a lot when comparing those having completed to those not having completed an upper secondary degree by age 21. Moreover, this holds true for all four Nordic countries under study. In several respects, there are also dis‐ tinct cross‐country differences among both completers and non‐ completers, as pointed out already in Chapter 2. Broadly speaking, about 90% of the completers are either continuing in education or working when aged 21, with this share being slightly higher in Denmark and marginally lower in Finland. The corresponding situation for 21‐year‐old non‐completers looks very different with a share in either full‐time education or employment well below 70% for Finland, Norway and Sweden. Only in Denmark is this share clearly higher (78%). From a within‐country perspective, these differences in combined education and employment shares imply that the “activity” gap between 21‐year‐old completers and non‐completers is compara‐ tively large for Sweden (26%‐points), closely followed by Norway and Finland, with Denmark coming out with the smallest but still quite nota‐ ble gap (15%‐points) (Table 6.1).


191

The share of full‐time students is by far highest (60%) among Nor‐ wegian completers while the share of completers in employment is highest – and of a similar size (about 46%) – in Denmark and Sweden. Conversely, the share of Danish and Swedish completers continuing in education when aged 21 is comparatively low, as is the share of 21‐year‐ old Norwegian completers having entered working life. A distinctly dif‐ ferent cross‐country pattern emerges for the non‐completers. In particu‐ lar, the highest share of 21‐year‐old non‐completers engaged in full‐time studies is observed for Denmark (40% compared to less than 30% in the other three countries). The situation is more or less the opposite when it comes to the non‐completers’ employment share: now Norway has the largest (close to 43%) and Denmark the lowest (below 38%) share. Figure 6.7: Main activities at age 21 for completers and non‐completers of an upper secondary education, based on pooled information on all three youth cohorts under study, by country

Completers 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Denmark Finland Norway Sweden Student Employed Unemployed Pensioner Other

192


Non‐completers 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Denmark Student

Finland

Employed

Norway

Unemployed

Pensioner

Sweden Other

Notes: Completers are defined as young people having completed an upper secondary education by age 21. Conversely, non‐completers are defined as those having reached 21 years‐of‐age without finishing an upper secondary degree. The number of completers is 103,203 for Denmark, 158,611 for Finland, 109,723 for Norway and 243,763 for Sweden, whereas the number of non‐completers is 61,676 for Denmark, 34,956 for Finland, 46,441 for Norway and 46,494 for Sweden. For definitions of the five main activity groups, see Chapter 2.

These highly different cross‐country patterns when it comes to both completers’ and non‐completers’ shares in education and employment when aged 21 also explain the huge variation observed across the four countries when contrasting the “activity” (education + employment) share of completers to that of non‐completers. As shown in Table 6.1, most of this “activity” gap originates in a dramatic cross‐country varia‐ tion in the gap between completers’ and non‐completers’ enrolment in education: the difference in the share of 21‐year‐old completers and non‐completers continuing in education is conspicuously large for Nor‐ way (an almost 35%‐point lower share for the non‐completers) com‐ pared to a gap of less than 6%‐points for Denmark. The corresponding gap in employment shares is typically much smaller (about 9%‐points in Denmark and 7%‐points in Sweden) or even reversed. Indeed, the em‐ ployment share is higher for 21‐year‐old non‐completers than for their completer peers in Finland (about 4%‐points higher) and especially in Norway (about 12%‐points higher).


193

Table 6.1: Main activities at age 21: completers (%‐share) vs. non‐completers (%‐point gap) Main activity

Student Employed Student + employed Unemployed Disability benefit (pensioner) Other (inactivity) NEET activities Total

Denmark

Finland

Norway

Sweden

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

45.9 46.6 92.5 2.3 0.0 5.2 7.5 100.0

‐5.8 ‐9.1 ‐14.9 +4.3 +1.8 +8.9 +15.0

53.5 34.9 88.4 7.8 0.3 3.5 11.6 100.0

‐24.6 +4.1 ‐20.5 +4.5 +3.8 +12.3 +20.6

60.4 30.3 90.7 2.6 0.2 6.5 9.3 100.0

‐34.6 +12.4 ‐22.2 +9.4 +3.9 +8.9 +22.2

43.2 46.0 89.2 7.0 0.2 3.5 10.7 100.0

‐18.8 ‐7.1 ‐25.9 +13.5 +8.3 +4.3 +26.1

Notes: These calculations are based on the distributions displayed in Figure 6.7. The %‐shares in the table show the distribution of completers across main activities when aged 21 (these shares are identical to those displayed in the upper graph of Figure 6.7). The %‐point gaps indicate the difference between non‐completers’ and completers’ shares. A negative sign implies that the non‐completers’ share is lower: for instance, the share of Danish non‐completers in education is 5.8 %‐points lower (or 40.1%) than the corresponding share for completers (45.9%). Likewise, a positive sign means that a larger share of the non‐completers than of the completers is in that particular activity.

The other side of the coin illustrates completers’ and non‐completers’ involvement in various NEET activities. The share of completers show‐ ing up in such activities when aged 21 is small in all four countries, which follows directly from most of them being either full‐time students or employed. Of the non‐completers, on the other hand, about one‐third is neither studying nor working by age 21. Only in Denmark is this share smaller, under one‐fourth. Moreover, a majority of the non‐completers in NEET activities belongs to the dumping category of “other” (inactivi‐ ty), that is, they do not appear in any of the large administrative regis‐ ters from which our national datasets are compiled. A conspicuous ex‐ ception from this pattern is Swedish non‐completers, though, whose NEET activities are dominated by registered unemployment. Notable shares of the non‐completers are also on disability benefits, a labour market outcome that is almost totally missing among 21‐year‐old com‐ pleters in all four countries. In view of these findings, it is hardly surprising that a comparison by country of completers’ and non‐completers’ NEET activities when aged 21 results in a situation where the non‐completers’ shares throughout exceed those of the completers. However, the cross‐country variation in also these patterns is substantial (Table 6.1). For instance, the share of the unemployed is quite high and of a similar size (about 7%) among Finnish and Swedish completers aged 21. The share of unemployed non‐ completers is, however, even larger (12.3% in Finland and 20.5% in Sweden). This results in a remarkably large gap (13.5%‐points) in Swe‐

194


dish completers’ and non‐completers’ unemployment shares, with the corresponding gap being notably lower (4.5%‐points) for Finland. A similar situation prevails for Denmark and Norway, where the share of 21‐year‐old completers in unemployment is very low (less than 3%) while their non‐completer peers experience either somewhat higher unemployment (Denmark) or much higher unemployment (Norway). As is evident in Table 6.1, similar examples can be found with respect to disability benefits, as well as other types of inactivity.

6.3.2

Situation ten years later, at age 26

Next we pick up these completers and non‐completers at age 26. As not‐ ed above, we thereby retain the classification of our young people into completers and non‐completers that prevailed when they turned 21. This exercise results in the country‐specific distributions across main activities shown in the two graphs contained in Figure 6.8. Among the completers, employment has by age 26 taken over as the overwhelmingly most common activity, while their share in full‐time education has shrunk to between one‐third (Denmark) and one‐fourth (Sweden). Taken together, the share of completers either studying or working is at this age slightly higher than five years earlier, at age 21, but only for Denmark and Sweden. For Finnish completers, the “activity” share stands at approximately the same level as five years earlier, obvi‐ ously mainly due to the conspicuous share outside both education and the labour market observed among young women at age 26 (and also at age 31) in the previous sub‐chapter. For Norway, in contrast, the share of completers either studying or working is slightly lower at age 26 than at age 21. This finding is in line with the observations made earlier in this chapter, viz. a decline in the share of students that exceeds the growth in the share in employment, coupled with a concomitant in‐ crease in the share of young people in inactivity (cf. Tables 6.1 and 6.2). Also the share of non‐completers in either education or employment is higher at age 26 than at age 21: marginally higher in Denmark (78 vs. 79%) and Norway (69 vs. 70%), but notably higher in Sweden (63 vs. 71%). Only for Finland do we see a decline in non‐completers’ “activity” share over these five years, just as for Finnish completers and, evidently, for the same main reason. In all four countries, the “activity” share of non‐ completers lags, nonetheless, far behind that of completers still at age 26 (Table 6.2). However, the cross‐country variation in this “activity” gap between completers and non‐completers is much smaller at age 26 than at age 21. In particular, while this activity gap between completers and non‐


195

completers has remained approximately unchanged from age 21 up to age 26 in Denmark, it has narrowed quite substantially in Sweden (cf. Tables 7.1 and 7.2). Yet, the way in which the gaps between completers and non‐ completers in student and employment shares feed into this “activity” gap looks totally different at age 26, when compared five years earlier. Note‐ worthy is especially the change in the employment gap: by age 26, it has turned from positive to negative (Finland and Norway) or even more neg‐ ative (Sweden). At age 26, non‐completers thus typically face a weaker (Norway) or much weaker (Finland and Sweden) employment situation than completers, except in Denmark (Table 6.2). The share of 21‐year‐old completers outside education and employment was noted to be low in all four countries. The share of completers in NEET activities has declined further by age 26. Among non‐completers of age 21, large shares were found to be either unemployed, on disability benefits or in other types of inactivity, with this share ranging from about 22% in Den‐ mark up to almost 37% in Sweden. The situation has, by age 26, improved slightly among Swedish non‐completers but only marginally so among Dan‐ ish and Norwegian non‐completers. For Finnish non‐completers, the NEET share is basically unchanged and now, in effect, highest among the four countries. However, just as the “activity” gap, also the NEET gap is con‐ structed in a different way at age 26 than age 21. Put differently, the gaps in completers’ and non‐completers’ shares in unemployment, disability ar‐ rangements and other inactivity feed into the total NEET gap in highly dif‐ ferent ways at age 21 and at age 26 (cf. Tables 6.1 and 6.2). On the whole, though, the overall cross‐country pattern in relation to completers’ and non‐completers’ distributions across main activities does not change that much when comparing the situation at age 26 to the situa‐ tion prevailing at age 21. The highest “activity” share is still observed for Denmark, and this holds true for both completers and non‐completers. The ranking of the other three countries has changed, though, with Sweden sur‐ passing both Finland and Norway with respect to both completers and non‐ completers. Then follows Norway and, finally, comes Finland. Indeed, at age 26, Finland comes out with the lowest “activity” share and, conversely, the highest NEET share among both completers and non‐completers. As noted above, this outcome for Finland seems to be due, at least in part, to the growing share beyond age 21 of young women outside both education and the labour market. Additionally, (registered) unemployment turns out to be quite widespread among Finnish non‐completers aged 26. However, as has been indicated in earlier parts of this report and as will be shown later on, this is largely a cohort effect related to the weak employment prospects faced by young labour market entrants in the 1990s.

196


Figure 6.8: Main activities at age 26 for completers and non‐completers of an upper secondary education, based on pooled information on two (1993 and 1998) of the youth cohorts under study, by country

Completers 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Denmark Student

Finland

Employed

Norway

Unemployed

Pensioner

Sweden Other


Finland

Employed

Norway

Unemployed

Pensioner

Sweden Other

Notes: Completers are defined as young people having completed an upper secondary education by age 21. Conversely, non‐completers are defined as those having reached 21 years‐of‐age without finishing an upper secondary degree. Recall that we are not able to trace the 2003 cohort of 16‐year‐olds up to age 26, only up to age 21. This also affects the total number of completers and non‐completers in respective country underlying the %‐shares displayed in this particular figure: now the number of completers is 68,003 for Denmark, 108,935 for Finland, 72,885 for Norway and 154,288 for Sweden, whereas the number of non‐completers is 36,762 for Denmark, 23,728 for Finland, 29,521 for Norway and 28,818 for Sweden. For definitions of the five main activity groups, see Chapter 2.


197



Denmark

Finland

Norway

Sweden

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

35.1 59.2 94.3 3.2 0.1 2.4 5.7 100.0

‐17.7 +2.7 ‐15.0 +4.9 +2.9 +7.2 +15.0

27.0 61.6 88.6 6.1 0.6 4.6 11.3 100.0

‐10.9 ‐10.2 ‐21.1 +9.2 +5.0 +6.9 +21.1

29.5 59.4 88.9 3.3 0.5 7.2 11.0 100.0

‐16.9 ‐2.2 ‐19.1 +8.5 +5.3 +5.5 +19.3

24.6 67.3 91.9 4.2 0.6 3.2 8.0 100.0

‐8.0 ‐12.5 ‐20.5 +7.8 +8.9 +3.9 +20.6

Notes: These calculations are based on the distributions displayed in Figure 6.8. The %‐shares in the table show the distribution of completers across main activities when aged 26 (these shares are identical to those displayed in the upper graph of Figure 6.8). The %‐point gaps indicate the difference between non‐completers’ and completers’ shares. A negative sign implies that the non‐completers’ share is lower: for instance, the share of Danish non‐completers in education is 17.7 %‐points lower (or 17.4%) than the corresponding share for completers (35.1%). Likewise, a positive sign means that the share of the non‐completers is larger than that of the completers in that particular activity.

6.3.3

Situation 15 years later, at age 31

Finally we compare the labour market outcomes of completers and non‐ completers at age 31, that is, 15 years after these young adults left com‐ pulsory education. At this particular age, about 95% of the Danish and 94% of the Swedish completers are either employed or enrolled in edu‐ cation (Figure 6.9, upper graph), that is, a still larger share than at age 26. In Finland, this “activity” share is 89%, which corresponds to an only marginally higher share than five years earlier. In Norway, on the other hand, this share has declined slightly, from 89% at age 26 to below 87% at age 31, a pattern observed also in earlier parts of this report. Of the Danish and Swedish completers, about 85% are in working life when aged 31, compared to some 76% of Finnish and Norwegian com‐ pleters. Hence, the completers’ employment situation has improved fur‐ ther by age 31.The share of completers still enrolled as full‐time stu‐ dents has shrunk to below 10% in Denmark and Sweden and to less than 11% in Norway. The highest share is found for Finland, or 13.5%.

198


While the employment situation has, in all four countries, improved also for the non‐completers over the 5‐year period from age 26 to age 31, the growth in the non‐completers’ employment share has been clear‐ ly weaker: up to a share of 73% in Denmark, 65% in Sweden, 62% in Norway and 60% in Finland (lower graph of Figure 6.9). As a conse‐ quence, the gap in employment shares between completers and non‐ completers is in all four countries larger at age 31 than at age 26 (cf. Table 6.2 and 6.3). Simultaneously, the gap in completers’ and non‐ completers’ student shares has turned small (Finland and Norway) or negligible (Denmark and Sweden). This also explains why the “activity” gap between completers and non‐completers is, in effect, slightly smaller at age 31 than at age 26, except for Sweden where it is unchanged and, hence, still highest among the four countries (Table 6.3). However, de‐ spite these various trends between and within countries up to age 31, the overall cross‐country pattern observed for completers and non‐ completers at age 26 shows up also five years later, at age 31: an em‐ ployment share and, hence, also an “activity” share of non‐completers lagging far behind that of completers. Figure 6.9: Main activities at age 31 for completers and non‐completers of an upper secondary education, based on information on one (1993) of the youth cohorts under study, by country

Completers 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Denmark Finland Norway Sweden Student Employed Unemployed Pensioner Other


199


Finland

Employed

Norway

Unemployed

Pensioner

Sweden Other

Notes: Completers are defined as young people having completed an upper secondary education by age 21. Conversely, non‐completers are defined as those having reached 21 years‐of‐age without finishing an upper secondary degree. Recall that we are not able to trace the 1998 and 2003 cohorts of 16‐year‐olds up to age 31. This affects the total number of completers and non‐completers in the respective country underlying the %‐shares displayed in this particular figure: now the number of completers is 36,932 for Denmark, 55,088 for Finland, 36,451 for Norway and 77,498 for Sweden, whereas the number of non‐completers is 19,778 for Denmark, 10,507 for Finland, 14,561 for Norway and 13,113 for Sweden. For definitions of the five main activity groups, see Chapter 2.

The steady increase up to age 31 in the share of Danish and Swedish completers in either education or employment is mirrored by a concom‐ itant decrease in their share in NEET activities. Indeed, the share of 31‐ year‐old completers outside both education and work is very low in Denmark and Sweden. In Finland, it is about 11% among both 26‐year‐ old and 31‐year‐old completers, with unemployment showing up as an important explanation for also 31‐year‐old completers’ comparatively high NEET share. In Norway, in contrast, the share of completers in NEET activities is slightly higher at age 31 (13%) than at age 26 (11%). This relatively large and increasing NEET share among Norwegian com‐ pleters seems to be mainly due to withdrawal not only from education but also from the labour market: the share of completers in unknown inactivity increases steadily from age 21 (6.5%) up to at age 31 (close to 10%). While a similar trend is discernible also for those receiving disa‐ bility benefits, the share of disability beneficiaries among Norwegian completers is still by age 31 below one per cent.

200




Denmark

Finland

Norway

Sweden

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

9.9 85.5 95.4 1.6 0.4 2.6 4.6 100.0

+0.9 ‐12.6 ‐11.7 +1.7 +4.0 +6.1 +11.8

13.5 75.5 89.0 4.8 1.1 5.1 11.0 100.0

‐2.9 ‐15.6 ‐18.5 +7.7 +5.8 +5.1 +18.6

10.7 76.0 86.7 2.8 0.9 9.7 13.4 100.0

‐3.6 ‐13.8 ‐17.4 +6.8 +6.2 +4.3 +17.3

9.0 85.3 94.3 2.1 1.4 2.3 5.8 100.0

+0.1 ‐20.6 ‐20.5 +5.8 +10.8 +3.8 +20.4

Notes: These calculations are based on the distributions displayed in Figure 6.9. The %‐shares in the table show the distribution of completers across main activities when aged 31 (these shares are identical to those displayed in the upper graph of Figure 6.9). The %‐point gaps indicate the difference between non‐completers’ and completers’ shares. A negative sign implies that the non‐completers’ share is lower: for instance, the share of Danish non‐completers in employment is 12.6 %‐points lower (or 72.9%) than the corresponding share for completers (85.5%). Likewise, a positive sign means that a larger share of the non‐completers than of the completers is in that particular activity.

Denmark shows up with the lowest NEET share (16.4%) also among 31‐ year‐old non‐completers, whereas substantially higher shares emerge for the other three countries: about 26% for Sweden, close to 30% for Finland and almost 31% for Norway. Hence, still by age 31 remarkably large shares of the non‐completers stand outside both education and employment. Indeed, their NEET share is only marginally lower than five years earlier, at age 26, and in Norway it reveals an increasing rather than decreasing trend. The moderate change in the non‐completers’ NEET share in adulthood is mainly explained by two opposite trends: declining shares of non‐completers in unemployment (compared with the situation at age 26) and growing shares of them withdrawing from the labour market, especially into disability arrangements. Accordingly it is not surprising that the gap in the share of completers and non‐ completers on disability benefits has increased in all four countries from age 26 up to age 31, whereas the corresponding gaps in unemployment and inactivity shares have typically declined over these five years.

6.3.4

The influence of economic shocks

As a final exercise of this sub‐chapter, we again split the information provided at age 26 into two parts: one illustrating the distribution across main activities at this particular age of young people belonging to the “economic‐bust” cohort of 1993, and one providing the same infor‐ mation for the “economic‐boom” cohort of 1998. Needless to say, the


201

focus now is on a comparison by cohort of the situation of completers vs. that of non‐completers. The outcome of this exercise is presented in Figure 6.10, separately for each of the four Nordic countries under study. The first graph of Figure 6.10 gives the distribution across main ac‐ tivities of Danish completers and non‐completers belonging to respec‐ tive youth cohort. The “activity” (education + employment) share of the 1998 economic‐boom cohort is slightly higher than for the 1993 eco‐ nomic‐bust cohort among both completers and non‐completers. Con‐ versely, the 1993 cohort is characterised by a somewhat higher NEET share irrespective of whether or not the young person has completed an upper secondary degree by age 21. In this sense, the 1998 economic‐ boom cohort seems to have experienced a more favourable labour mar‐ ket situation when aged 26. Simultaneously, however, the changes across NEET activities have been remarkable, and not necessarily always to the favour of the 1998 cohort. In particular, while the share in regis‐ tered unemployment is smaller (completers) or dramatically smaller (non‐completers) in the 1998 cohort, when compared to the 1993 co‐ hort, the situation is reversed when it comes to time spent outside both education and the labour force: higher shares of young people belonging to the 1998 cohort are, when aged 26, either on disability benefits or in other (unknown) types of inactivity. Moreover, this holds true for both completers and non‐completers. These findings point to a “reshuffling” of NEETs rather than to a straightforward impact of an economic shock. The next graph provides the corresponding information for Finland. As for Denmark, we see a clear improvement of completers’ and espe‐ cially of non‐completers’ engagement in education and employment across the two cohort, and a corresponding reduction in their NEET ac‐ tivities. But in contrast to Denmark, the shares among both completers and non‐completers in disability arrangements or other types of inactiv‐ ity are basically the same in the two cohorts. In other words, the decline in NEET activities across the two cohorts is in Finland almost entirely explained by less registered unemployment in the 1998 economic‐boom cohort than in the 1993 economic‐bust cohort. These findings are well in line with the high unemployment levels that prevailed long after the start of the economic recovery in 1994.

202


Figure 6.10: Distribution (%‐share) of completers and non‐completers of an upper secondary education across main activities at age 26, by country; comparison of the “economic‐bust” cohort of 1993 with the “economic‐boom” cohort of 1998

Denmark 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Completers, Completers, Non‐completers, Non‐completers, cohort 1993 cohort 1998 cohort 1993 cohort 1998 Student Employed Unemployed Pensioner Other Finland 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Completers, Completers, Non‐completers, Non‐completers, cohort 1993 cohort 1998 cohort 1993 cohort 1998 Student Employed Unemployed Pensioner Other


203

Norway 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Completers, cohort 1993

Student

Completers, cohort 1998

Employed

Non‐completers, Non‐completers, cohort 1993 cohort 1998

Unemployed

Pensioner

Other

Sweden 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Completers, cohort 1993

Student

Completers, cohort 1998

Employed

Non‐completers, Non‐completers, cohort 1993 cohort 1998

Unemployed

Pensioner

Other

Notes: See Figure 6.7.

The graph displaying the situation for Norwegian completers and non‐ completers belonging to either one of the two cohorts repeats a pattern that has already been discernible in previous parts of this report: a drop in the share of students that is not fully compensated by the concomitant increase in employment, resulting in an “activity” (education + employ‐ ment) share that reveals a declining rather than increasing trend over cohorts (as well as within cohorts with age). Simultaneously, we see more young people moving outside both education and the labour force. This trend shows up strongly among both completers and non‐completers.

204


The last graph of Figure 6.10 illustrates the situation for Sweden. The combined education and employment share is slightly higher for com‐ pleters belonging to the 1998 economic‐boom cohort, but among non‐ completers it is of approximately of the same size in the two cohorts. Hence, while the completers’ NEET share is slightly lower in the 1998 cohort, it is unchanged across the two cohorts for non‐completers. As in Denmark and Norway, much more seems to happen with the composi‐ tion of NEETs: lower shares of unemployed jobseekers in the 1998 co‐ hort but at the expense of higher shares outside both education and the labour force. Again, the same pattern shows up among both completers and non‐completers. All in all, the most profound difference between completers and non‐ completers is much weaker employment prospects of non‐completers and, hence, notably higher risks of ending up in NEET activities. Eco‐ nomic business cycles, on the other hand, seem to impact on young peo‐ ple in much the same way irrespective of their educational background (completion or not of an upper secondary degree by age 21). But with an initially weaker labour market attachment of non‐completers, it is not surprising that also the consequences of an economic shock are typically conceived to be more serious for low‐skilled young people. Apart from a persistently higher NEET share among non‐completers, another con‐ spicuous trend that seems to be little, if at all, linked to the prevailing economic situation is the growing tendency among non‐completers of moving into potentially more risky NEET activities, that is, withdrawal not only from education but also from the labour force, including regis‐ tered unemployment. In Norway, this trend is for some reason conspic‐ uously common and increasing also among completers.

6.3.5

Main findings

The main finding is, unsurprisingly, that lower shares of non‐completers than of completers are either in education or in employment and that this gap prevails up to age 31, at least. In other words, non‐completers do not close up the gap to completers in their “activity” (education + employ‐ ment) share. Simultaneously, however, this gap to completers in terms of education and employment seems to be relatively small, or at least smaller than would perhaps have been expected in view of the discussion of low‐ skilled young people being the losers in today’s labour markets. Indeed, the conspicuous “success rates” in terms of studying and working observed for also non‐completers could be taken to indicate that, in the last resort, many of them fare reasonably well in the labour


205

market as young adults despite a low formal education. The highest suc‐ cess rates are obtained for Denmark. The difference in non‐completer outcomes between Denmark, on the one hand, and Finland and Sweden, on the other hand, could then be interpreted as a result of notable cross‐ country differences in the composition of the group of non‐completers. The so‐called hard core of non‐completers, that is, young people with disproportionally weak labour market prospects, tend to drop out from education at an early age in principally any country. If the number of non‐completers increases, this most likely implies that also young peo‐ ple with less serious problems and, hence, with an obviously closer la‐ bour market attachment, are for some reason shifting into the group of non‐completers. If this is the case, then the overall size of the group of non‐completers could also tell us something about the composition of non‐completers. For this very reason we would expect the relatively large share of 21‐year‐old Danish non‐completers to do better on aver‐ age than the comparatively small shares of non‐completers among the 21‐year‐olds observed for Finland and Sweden (cf. Chapter 2). However, this interpretation does not get support when comparing Finland and Sweden to Norway. In this setting, a large cross‐country difference in non‐completion rates does not result in conspicuously dif‐ ferent success rates of the three countries’ non‐completers. Moreover, even in the case of relatively high employment rates also among non‐ completers, previous research has shown that there is a large and signif‐ icantly negative wage differential between employed non‐completers and employed completers (see Bratsberg et al., 2010). Apart from differences in the quality of completers’ and non‐ completers’ employment contracts, a dimension overlooked in our anal‐ yses, there are worrying aspects also related to the time trend in non‐ completers’ labour market prospects. In particular, the results presented in this sub‐chapter point to the employment share of non‐completers lagging increasingly behind that of completers when comparing their situation at three different age points (21, 26 and 31). This is highly evi‐ dent in Table 6.4, which reproduces parts of the information provided in Tables 6.1 to 6.3. Simultaneously, the cross‐country variation in the gaps having emerged by age 31 between completers’ and non‐completers’ employment shares is particularly pronounced when it comes to Den‐ mark and Sweden: in both countries, completers fair equally well in terms of employment, whereas the gap to the non‐completers’ employ‐ ment share is almost twice as large in Sweden (20.6 %‐points) as it is in Denmark (12.6 %‐points).

206


Table 6.4: Selected main activities at age 21, 26 and 31, respectively: completers (%‐share) vs. non‐completers (%‐point gap) Main activity

Denmark

Finland

Norway

Sweden

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

%‐ share

%‐point gap

Employed at age 21 at age 26 at age 31

46.6 59.2 85.5

‐9.1 +2.7 ‐12.6

34.9 61.6 75.5

+4.1 ‐10.2 ‐15.6

30.3 59.4 76.0

+12.4 ‐2.2 ‐13.8

46.0 67.3 85.3

‐7.1 ‐12.5 ‐20.6

Disability benefit (pensioner) at age 21 at age 26 at age 31

0.0 0.1 0.4

+1.8 +2.9 +4.0

0.3 0.6 1.1

+3.8 +5.0 +5.8

0.2 0.5 0.9

+3.9 +5.3 +6.2

0.2 0.6 1.4

+8.3 +8.9 +10.8

Other (inactivity) at age 21 at age 26 at age 31

5.2 2.4 2.6

+8.9 +7.2 +6.1

3.5 4.6 5.1

+12.3 +6.9 +5.1

6.5 7.2 9.7

+8.9 +5.5 +4.3

3.5 3.2 2.3

+4.3 +3.9 +3.8

Notes: See Tables 6.1 to 6.3.

The counterpart to this development is a steadily growing share of non‐ completers moving outside both education and the labour market, either into disability arrangements or other (unknown) types of inactivity. While the share of disability beneficiaries increases with age also among young completers, this share is initially much higher and also increases much faster among the non‐completers, a pattern repeated for all four countries. When it comes to other types of inactivity, the situation varies substan‐ tially across countries, as also pointed out earlier. For all four countries, we observe a narrowing gap in completers’ and non‐completers’ “other” inactivity shares (Table 6.4). However, in both Denmark and Sweden this is due to a decline with age in the inactivity share among both completers and non‐completers, with this trend being stronger for the non‐ completers. In Finland and especially in Norway, on the other hand, we see growing numbers of completers withdrawing from both education and the labour force when going from age 21 up to age 31, whereas an opposite trend is observable among non‐completers. Indeed, the share in unknown inactivity among 31‐year‐old Norwegian completers is not only much higher than the corresponding share for completers in the other three countries. It also exceeds by far the inactivity share of both Danish and Swedish non‐completers aged 31, while it is only marginally lower than the inactivity share of 31‐year‐old non‐completers in Finland. Moreover, these overall patterns and trends across and within the four Nordic countries under study, as well as across and within the three youth cohorts under scrutiny, seem to have been only marginally, if at all, affected by changes in the economic environment. Changing business cycles seem to typically impact on young people in much the same way


207

irrespective of their upper secondary graduation background (complet‐ er vs. non‐completer). However, with an initially much weaker labour market attachment, the consequences are inevitably more far‐reaching for young non‐completers than for their completer peers. Finally, we undertake a rescaling of also these findings by relating them to the full youth population of each of the four Nordic countries under study. Table 6.5 displays the outcome of this exercise by main activity at age 21, 26 and 31, respectively, separately for completers and non‐completers. About 50% of both Finnish and Norwegian youth are enrolled in full‐ time education when aged 21, with completers contributing overwhelm‐ ingly to this high share. Sweden, in turn, comes out with the lowest share of students at this particular age mainly due to a very small share (under 4%) of young non‐completers continuing in education when aged 21. For Denmark, in contrast, we observe a strikingly high share of 21‐year‐ old non‐completers still in education. Conversely, Sweden turns out to have the highest share of 21‐year‐olds in employment, but this holds true for completers only. Young Swedes with only an exam from com‐ pulsory school still when aged 21 are the least likely of young Nordic non‐completers of being in education or employment. Table 6.5: Distribution of young completers and non‐completers by main activity at age 21, 26 and 31, respectively, by country, %‐share of the full youth population Status at age 21 – %‐share of the full population Completers Main activity Student Employed Student + employed Unemployed Pensioner Other NEET share In total

Non‐completers

DK

FI

NO

SW

28.8 29.2 58.0 1.4 0.0 3.3 4.7 62.7

43.9 28.6 72.5 6.4 0.2 2.9 9.5 82.0

42.5 21.3 63.8 1.8 0.1 4.6 6.5 70.3

36.3 38.6 74.9 5.9 0.2 2.9 9.0 84.0

DK

FI

NO

SW

15.0 14.0 29.0 2.5 0.7 5.3 8.5 37.3

5.2 7.0 12.2 2.2 0.7 2.8 5.7 18.0

7.7 12.7 20.4 3.6 1.2 4.6 9.4 29.7

3.9 6.2 10.1 3.3 1.4 1.2 5.9 16.0

Status at age 26 – %‐share of the full population Completers Main activity Student Employed Student + employment Unemployed Pensioner Other NEET share In total

208

Non‐completers

DK

FI

NO

SW

22.0 37.1 59.1 2.0 0.1 1.5 3.6 62.7

22.1 50.5 72.6 5.0 0.5 3.8 9.3 82.0

20.7 41.8 62.5 2.3 0.4 5.1 7.8 70.3

20.7 56.5 77.2 3.5 0.5 2.7 6.7 84.0


DK

FI

NO

SW

15.0 14.0 29.0 2.5 0.7 5.3 8.5 37.3

2.9 9.3 12.2 2.8 1.0 2.1 5.9 18.0

3.7 17.0 20.7 3.5 1.7 3.8 9.0 29.7

2.7 8.8 11.5 1.9 1.5 1.1 4.5 16.0

Status at age 31 – %‐share of the full population

Completers

Main activity Student Employed Student + employed Unemployed Pensioner Other NEET share In total

DK

FI

NO

6.2 53.6 59.8 1.0 0.3 1.6 2.9 62.7

11.1 61.9 73.0 3.9 0.9 4.2 9.0 82.0

7.5 53.4 60.9 2.0 0.6 6.8 9.4 70.3

SW 7.6 71.7 79.3 1.8 1.2 1.9 4.9 84.0

Non‐completers DK

FI

NO

SW

4.0 27.2 31.2 1.2 1.6 3.2 6.0 37.3

1.9 10.8 12.7 2.3 1.2 1.8 5.3 18.0

2.1 18.5 20.6 2.9 2.1 4.2 9.2 29.7

1.5 10.4 11.9 1.3 2.0 1.0 4.3 16.0

Notes: See Tables 6.1 to 6.3.

Sweden, together with Finland, also stands out with a comparatively high share of 21‐year‐olds registered as unemployed jobseekers, but only among young completers. Among 21‐year‐old non‐completers, Norway shows up with the highest unemployment ratio, albeit the dif‐ ference to the other three countries is rather small. However, together with Denmark, Norway also has a relatively high share of 21‐year‐olds whose activity is unknown. Moreover, this high inactivity share shows up among both young completers and young non‐completers. By age 26, Sweden still comes out with the highest employment share in the youth population but, again, only for completers; Swedish non‐ completers’ employment situation is weakest among Nordic non‐ completers also at age 26. The share of young people enrolled in full‐ time education still when aged 26 has, in turn, converged across the four countries and, moreover, among both completers and non‐completers. The only exception is Denmark, where the share of non‐completers con‐ tinuing as full‐time students is at the same high level as five years earli‐ er, at age 21. When it comes to NEET activities, the overall cross‐country pattern at age 26 is very similar to the situation observed at age 21. In particular, the share of the unemployed has remained relatively high in both Fin‐ land and Sweden due to unemployment still being more common among Finnish and Swedish completers, when compared to Danish and Norwe‐ gian completers. And also among the 26‐year‐olds, Norway comes out with the highest share of unemployed jobseekers among the non‐ completers. Also the share of young Norwegians outside both education and the labour market is comparatively high still at age 26, with the in‐ activity share being even higher among completers than among non‐ completers. While the inactivity share has remained at a high level also among Danish non‐completers, the situation has improved among Dan‐ ish completers.


209

By age 31, finally, all four countries have around 10% of their youths enrolled as full‐time students, with the student share among the non‐ completers now being very low. The employment share is high among Finnish and especially among Swedish completers, whereas the em‐ ployment situation of the countries’ non‐completers is relatively weak still at age 31. A reversed situation is observed for Denmark and Norway in the sense that the employment share is comparatively low among completers but relatively high among non‐completers. The share of young people in unemployment has come down in all four countries and, moreover, among both completers and non‐completers but is, nonetheless, still highest among Norwegian non‐completers. Likewise, Norway has the highest share of young people outside both education and the labour market also among the 31‐year‐olds. Indeed, compared to the situation at age 26, the inactivity share has increased rather than de‐ creased among both completers and non‐completers. On the whole, then, the overall impression from this exercise of rescal‐ ing the main activity shares at three age points of completers and non‐ completers in relation to the full youth population of each country, does not seem to change the general picture mediated so far. Particularly strik‐ ing is the stability with age in the NEET share among the non‐completers. Another conspicuous feature of Table 6.5 is the similarity in NEET shares across the four countries, a similarity that is higher than expected in view of the large differences observed in school‐to‐work‐transition patterns and completing‐rate time‐profiles between the countries.

6.4 Main activities beyond age 21 – late completion vs. non‐completion We end this chapter on labour market outcomes in adulthood by relax‐ ing our “prime” definition of non‐completers. More precisely, we now allow for the fact that some of the non‐completers, but far from all, do finalise an upper secondary education, but only after the age of 21, i.e., more than five years after having left compulsory school. As shown ear‐ lier, notably in Chapter 2, late completion of upper secondary education is common among young people in Denmark and Norway. In Finland and Sweden, on the other hand, a majority of young people has finalised an upper secondary education by the time they turn 21. We start this sub‐chapter by exploring to what extent young people with only a basic education still at age 21 succeed in completing an up‐ per secondary education later in life. Again, this is done for two age

210


points: by the time they turn 26 and by the time they turn 31. In the fol‐ lowing, these young people are called “late completers” in order to sepa‐ rate them from 21‐year‐old completers, that is, young or early complet‐ ers. Likewise, we use the term “adult non‐completer” for young people lacking an upper secondary degree still at age 26 or age 31 in order to separate them from 21‐year‐old non‐completers, that is, young non‐ completers. In a next step, we compare the labour market outcomes of young completers, late completers and adult non‐completers at these same age points. This particular focus on late completers implies that most attention is paid to young people identified as non‐completers at age 21. More pre‐ cisely, the emphasis is on whether or not these 21‐year‐old non‐ completers succeed in achieving an upper secondary certificate by age 26 or by age 31, and to what extent this late completion eventually af‐ fects their labour market situation when compared to early completers and, especially, to adult non‐completers. In other words, is early comple‐ tion the best choice, late completion a second‐best choice and non‐ completion still in adulthood the worst alternative in terms of labour market outcomes as a young adult? This setting also means that all sub‐ sequent results are based on information on only two out of our three youth cohorts for the simple reason that the youngest cohort (the 2003 cohort of 16‐year‐olds) cannot be traced beyond the age of 21.

6.4.1

Late completion of upper secondary education

There might be reason to first recall the shares of 21‐year‐old non‐ completers in the four Nordic countries under study. We therefore start by reproducing the information provided in Table 2.2 of Chapter 2, but now with the youngest (2003) cohort of 16‐year‐olds left out. As has been pointed out earlier in this report, non‐completion of an upper sec‐ ondary degree is strikingly common among 21‐year‐old Danes, less so among 21‐year‐old Norwegians and quite infrequent among 21‐year‐old Finns and Swedes. Table 6.6: Non‐completion rates (%‐shares) at age 21 in four Nordic countries, by cohort Cohort 16‐year‐olds in 1993 16‐year‐olds in 1998

Denmark

Finland

Norway

Sweden

34.7 39.0

16.0 19.7

28.5 29.1

14.5 17.0

The next table (Table 6.7) shows to what extent these 21‐year‐old non‐ completers have succeeded in finalising an upper secondary degree ei‐ ther by age 26 or by age 31. In Denmark, more than one‐half of the 21‐


211

year‐old non‐completers from the 1993 cohort had eventually complet‐ ed an upper secondary degree by the time they turned 31. Close to 42% of them had actually come around to completing an upper secondary education by age 26. This high share of late completers also explains the rapid decline beyond age 21 in the non‐completion share displayed in Figure 2.2a of Chapter 2 for the oldest (1993) Danish cohort of 16‐year‐ olds. Moreover, this high share of late completers is repeated for the Danish 1998 cohort (about 43% by age 26). Table 6.7: Completion of an upper secondary degree by age 26 and by age 31 among 21‐year‐old non‐completers, by country

Completion by age

Denmark

Cohort 16‐year‐olds in 1993 16‐year‐olds in 1998

Finland

26

31

26

31

41.6 42.9

51.8

25.4 29.7

34.7

Norway 26

Sweden

31

26

31

26.3 36.3 29.1

19.3 16.7

26.4

In the other three countries, the completion rates beyond age 21 are much lower. In Finland and Norway, slightly more than one‐third of the young people in the 1993 cohort identified as non‐completers still at age 21 had achieved an upper secondary certificate by the time they turned 31 with the corresponding share being only about 26% for Sweden. Moreover, while the late‐completion rate reveals an increasing trend across cohorts in both Finland and Norway, it has rather been declining in Sweden. These cross‐Nordic differences in young non‐completers’ likelihood of achieving an upper secondary certificate only in adulthood are quite pronounced. It could be argued that the substantially lower non‐ completion rates beyond age 21 observed for Finland, Norway and Swe‐ den, when compared to Denmark, are related to the comparatively high share of non‐completers among 21‐year‐old Danes (Table 6.6). However, the markedly higher non‐completion share among 21‐year‐old Danes can only explain part of these conspicuous cross‐country differences in late‐completion rates. In particular, while the late‐completion rates are very similar for Finland and Norway, the share of non‐completers by age 21 is highly different in the two countries. Likewise, while the share of 21‐year‐old non‐completers is of similar size in Finland and Sweden, there are substantial differences between the two countries when it comes to late‐completion rates. In other words, there is no clear‐cut cross‐country correlation between the early‐ and late‐completion rates reported in Tables 6.6 and 6.7.

212


6.4.2

Late completers: comparison of labour market outcomes at age 26

The completion patterns of post‐compulsory‐school degrees displayed in Tables 6.6 and 6.7 raise several questions: Do early completers tend to fair better in terms of labour market outcomes than late completers? Likewise, do late completers typically fair better than adult non‐ completers? The next figure aims to provide at least part of an answer to these questions with the focus being on the situation prevailing among 26‐year‐olds differing in their completion and non‐completion history. In order to clean the picture from changing cohort‐specific distributions across main activities as well as from business cycle fluctuations, Figure 6.11 is based on information for the 1998 cohort only, a cohort of young people that can be traced up to age 26, but not to age 31. We return to the situation among 31‐year‐olds in the next section. Figure 6.11 contains four country‐specific graphs, each of which dis‐ plays three distributions across our five main activity categories. All three distributions refer to the situation prevailing at age 26, whereas the allocation of these 26‐year‐olds across the three distributions de‐ pends on their completion and non‐completion history with respect to post‐compulsory‐school educations. The first (left‐hand‐side) pillar il‐ lustrates the distribution across main activities of the 1998 cohort’s young completers when aged 26, that is, of those young people in the cohort who had completed an upper secondary education already by age 21. The second (middle pillar) shows the corresponding distribution for the cohort’s late completers, that is, those young people in the cohort who were classified as non‐completers when aged 21 but who had suc‐ ceeded in completing an upper secondary degree by age 26. The third (right‐hand‐side) pillar, finally, displays the distribution across main activities of the cohort’s adult non‐completers, that is, those young peo‐ ple in the cohort who still when aged 26 had an exam only from compul‐ sory school. The first graph of Figure 6.11 illustrates the situation for Danish young completers, late completers and adult non‐completers belonging to the 1998 cohort. The differences in main‐activity distributions be‐ tween young completers and late completers are marginal at age 26. In both groups, engagement in either education or employment is the overwhelmingly most common activity. Hence, the advantage of com‐ pleting an upper secondary degree by age 21 instead of age 26 seems to be minor in Denmark, at least in terms of labour market outcomes in adulthood. The situation is entirely different for Danish adult non‐ completers: an “activity” (education + employment) share of about 74%


213

implies that one‐fourth of the adult non‐completers is in NEET activities when aged 26, belonging mainly to the dumping category of “other” (un‐ known) inactivity. The next graph gives the corresponding information for Finland. It clearly shows that young completers fare much better than late com‐ pleters. In particular, about nine out of ten young completers are en‐ gaged in either education or employment by age 26. Among late com‐ pleters, this share is approximately ten percentage points lower, or some 80%. Instead, they face a higher risk of showing up as unemployed jobseekers or disability beneficiaries, when compared to young com‐ pleters. However, compared to adult non‐completers, these late com‐ pleters do quite well. The “activity” share of adult non‐completers is only about 65%. While they have an only marginally higher risk of becoming unemployed, when compared to late completers, their risk of ending up in either disability arrangements or other types of inactivity is consider‐ ably higher than for late completers. Hence, early completion, late com‐ pletion and non‐completion still as a young adult makes a considerable difference in Finland. A more or less similar overall pattern emerges for Norway except that the main‐activity distributions of young and late completers are more similar in Norway than in Finland. Indeed, the difference in em‐ ployment shares between young and late completers is marginal in Norway, whereas there are larger differences between the two groups when it comes to enrolment in education and registered unemployment: the share enrolled in education still at age 26 is higher among young completers, whereas the share in registered unemployment is higher among late completers. A conspicuous finding, however, is that the dif‐ ference in inactivity shares among young and late completers is close to negligible, implying that young and late completers are equally likely to withdraw outside both education and the labour market. The distribu‐ tion across main activities of adult non‐completers is distinctly less fa‐ vourable and, in effect, remarkably similar to that observed for adult non‐completers in Finland: an “activity” share just above 60% in combi‐ nation with large shares being in NEET activities already at age 26.

214


Figure 6.11: Main‐activity distributions (%‐share) by age 26 of young complet‐ ers vs. young non‐completers having completed (late completer) or not having completed (adult non‐completer) an upper secondary education by age 26, based on information for the 1998 youth cohort, by country

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completers by age 21 Student Employed 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completers by age 21 Student Employed

Denmark

Late completers by age 26

Unemployed

Adult non‐completers by age 26

Pensioner

Other

Finland


Unemployed


Pensioner

Other


215

Norway 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completers by age 21

Student

Employed


Unemployed


Pensioner

Other

Sweden 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completers by age 21

Student

Employed


Unemployed


Pensioner

Other

Notes: For definitions of the five main activity groups, see Chapter 2. Young completers = young people having completed an upper secondary education already by age 21. Late completers = 21‐ year‐old non‐completers having achieved an upper secondary degree by age 26 (but not yet by age 21). Adult non‐completers = young people with only a basic education still at age 26.

A third kind of pattern is discernible for Sweden (last graph of Figure 6.11). As noted earlier in this report, young Swedish completers fare ex‐ tremely well: their “activity” (education + employment) share exceeds 90% by age 26, a share that lags only slightly behind that of young Danish completers. Late completers, in contrast, do not seem to manage better in terms of labour market outcomes than do adult non‐completers: the dif‐ ference in main‐activity distributions is strikingly small. While the “activi‐ ty” share of 26‐year‐old non‐completers is about 68%, it is only slightly

216


higher for late completers. Hence, when compared to the other three countries, Swedish late completers fare worst, whereas Swedish adult non‐completers fare better than both Finnish and Norwegian adult non‐ completers, and only slightly worse than Danish adult non‐completers.

6.4.3

Late completers: comparison of labour market outcomes at age 31

Next we compare the labour market outcomes of late completers and adult non‐completers five years later, at age 31 (Figure 6.12). More pre‐ cisely, we amend the setting used in the previous figure and add one more distribution: a pillar showing the distribution across main activities of young non‐completers having achieved an upper secondary degree only by age 31 (but not yet by age 26). In other words, Figure 6.12 displays the activity distribution at age 31 for a total of four different categories of young people: those who had completed an upper secondary degree al‐ ready by age 21 (young completer); those who had completed an upper secondary degree either by age 26 (late completers by 26) or by age 31 (late completer by 31); and those having an exam only from compulsory school still when aged 31 (adult non‐completer). Consequently, all four distributions displayed in Figure 6.12 cover merely the 1993 cohort, as the two younger cohorts cannot be traced up to age 31. The first graph of Figure 6.12, displaying the situation at age 31 for young Danes differing in their completion and non‐completion history, reveals several interesting results. First, the pattern of no clear‐cut dis‐ advantage from completing an upper secondary degree only by age 26, instead of age 21, is repeated also for the 1993 cohort. Second, comple‐ tion only by age 31 seems to have been an even better option, at least in terms of labour market outcomes. Third, adult non‐completion, now at age 31, results in much the same outcome as adult non‐completion at age 26: an “activity” share of about 74% and one‐fourth of adult non‐ completers in NEET activities. Finally, when comparing the results for young completers and late completers by age 26 with those displayed in Figure 6.11 for the 1998 cohort, the 1993 cohort seems to have fared slightly worse in terms of unemployment, which is in line with the re‐ sults reported earlier in this chapter (when comparing the 1993 “eco‐ nomic‐bust” cohort to the 1998 “economic‐boom” cohort). The conspic‐ uously good labour market performance of young people having gradu‐ ated from upper secondary education only by age 31 may, in effect, also be an indication of the influence of these deep economic recession years (cf. the discussion in the introductory chapter).


217

Figure 6.12: Main activity distributions (%‐share) by age 31 of young completers vs. young non‐completers having completed an upper secondary degree by age 26 (late completers by 26) or only by age 31 (late completers by 31), or not hav‐ ing completed an upper secondary education by age 31 (adult non‐completers), based on information on the 1993 youth cohort, by country

Denmark 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completer Late completer Late completer Adult non‐ by age 21 by age 26 by age 31 completer by age 31 Student Employed Unemployed Pensioner Other Finland 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completer Late completer Late completer Adult non‐ by age 21 by age 26 by age 31 completer by age 31

218

Student

Employed

Unemployed


Pensioner

Other

Norway 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Young completer by age 21

Student

Late completer by age 26

Employed


Unemployed

Pensioner

Adult non‐ completer by age 31

Other

Sweden 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Young completer by age 21

Student


Employed


Unemployed

Pensioner

Adult non‐ completer by age 31

Other

Notes: For definitions of the five main activity groups, see Chapter 2. Young completers = young people having completed an upper secondary education already by age 21. Late completers by 26 = 21‐year‐old non‐completers having achieved an upper secondary degree by age 26 (but not yet by age 21). Late completers by 31 = 21‐year‐old non‐completers having achieved an upper secondary degree by age 31 (but not yet by age 26). Adult non‐completers = young people with only a basic education.

The second graph of Figure 6.12, showing the corresponding results for Finland, paints a slightly different picture, despite certain distinct fea‐ tures which Finland seems to share with Denmark. In particular, young completers are most successful with respect to labour market outcomes, followed by those having completed an upper secondary degree only by age 31. The lowest “activity” (education + employment) share is ob‐


219

tained for those having graduated by age 26 which, again, may well re‐ flect the impact of the economic recession in the early 1990s. Another sign of the influence of these recession years is the higher shares of un‐ employed in the 1993 than in the 1998 cohort. Adult non‐completers, on the other hand, are also in Finland distributed across main activities in much the same way as adult non‐completers in the 1998 cohort: both reach up to an “activity” level of only some 65%. The share of the unem‐ ployed is in both cohorts about 14%, whereas about one‐fifth has with‐ drawn outside the labour market. The pattern emerging for Norway, as displayed in the third graph of Figure 6.12, indicates that the two groups of late completers of an up‐ per secondary degree fare, by and large, equally well in terms of com‐ bined education and employment shares, but worse than young com‐ pleters. There is, however, one major difference between the two groups of late completers: a relatively larger share of unemployed among those having graduated by age 26 and a relatively larger share having withdrawn from both education and the labour market among those having graduated only by age 31. The weakest labour market situation is observed for adult non‐completers. Indeed, their distribu‐ tion across main activities is very similar to that of adult non‐ completers in the 1998 cohort also in Norway. The last graph of Figure 6.12 explores the situation at age 31 among young Swedes belonging to the 1993 cohort. Of the four groups of young people differing in their graduation history, only the young completers reach an “activity” share of almost 92% when aged 31. This share is marginally lower than that of young completers in the 1998 cohort, mainly due to a slightly higher share of unemployed jobseekers among young completers in the 1993 cohort. The outcome for the other three groups of young people is notably weaker with an “activity” share of about 70%, irrespective of late graduation or no graduation. Indeed, the main difference between the three groups of late completers and adult non‐completers is not in their combined education and employment share but rather in the composition of their NEET activities: those hav‐ ing graduated by age 26 experience, when 31‐year‐old, more unem‐ ployment but less disability, with the situation being the opposite for those having graduated only by age 31 as well as for those having no post‐compulsory degree still at age 31. Perhaps most striking is the steady increase in the share of disability beneficiaries and also of those in other types of inactivity when going from young completers to late completers to adult non‐completers.

220


6.4.4

Main findings

This sub‐chapter has focused on exploring whether there are distinct differences in terms of labour market outcomes between young com‐ pleters, late completers and adult non‐completers of an upper secondary degree. This comparison was made at two age points: among those aged 26 and among those aged 31. The results obtained for young people when aged 26 indicate the fol‐ lowing. We can identify basically three patterns characterising the four Nordic countries under study. In Denmark, early and late completion of an upper secondary degree seems to make no difference in terms of labour market outcomes. Indeed, the distributions across main activities are almost identical for young completers and late completers, with most of them either studying or working at age 26. Adult non‐ completers, on the other hand, are less likely to continue in education. They also face weaker employment possibilities and, consequently, a higher risk of ending up in NEET activities. A second kind of pattern emerges for Finland and Norway. In both countries, young completers fare better than late completers, albeit the difference in labour market outcomes is more pronounced in Finland than in Norway. Conversely, the gap to adult non‐completers is broader in Norway than in Finland. Indeed, the labour market situation of 26‐ year‐old non‐completers is strikingly similar for these two countries, with almost 40% of them being in NEET activities. Sweden, finally, is characterised by a third type of pattern. In particu‐ lar, while young completers fare extremely well in terms of labour mar‐ ket outcomes, late completers lag far behind. A conspicuous finding is that there are only small, if any, differences in the distributions across main activities for late completers and adult non‐completers. In both groups, about 30% are in NEET activities when aged 26. Hence, adult non‐completers do not seem to fare worse than late completers, in terms of labour market outcomes. Late completers seem to be the major loser group in this setting. Taken together, these highly different young‐completer/late‐ completer/adult‐non‐completer patterns observed for Denmark, Fin‐ land, Norway and Sweden imply that late completers and adult non‐ completers encounter highly different labour market situations in the four countries. Among late completers, Swedish late completers fare worst, whereas Danish late completers manage equally well as young completers in terms of labour market outcomes by age 26. The ranking of countries changes slightly when comparing the position of adult non‐ completers: Denmark shows up with the highest “activity” (education +


221

employment) share also among adult non‐completers, but now with Sweden coming second. Hence, while adult non‐completers throughout fare much worse than young completers in terms of labour market out‐ comes in adulthood it is not obvious that late completion is always the second‐best option. In particular, late completion can definitively be seen as a second‐best option in Finland and Norway, but not necessarily in Denmark and Sweden, albeit for different reasons: in Denmark late completers fare equally well as young completers, whereas in Sweden late completers fare no better than adult non‐completers. When extending this exercise with five more years, up to age 31, the cross‐country pattern becomes more dispersed. An additional circum‐ stance affecting the pattern within as well as between countries relates to the economic recession in the early 1990s, as the results for the 31‐ year‐olds concern the 1993 cohort only. This impact shows up in, for instance, typically higher shares of unemployed among both early and late completers in the 1993 cohort, when compared to the 1998 cohort. This finding thereby also lends further support to the conclusion drawn earlier in this chapter in relation to the impact of economic shocks on young completers and non‐completers; i.e., changing business cycles seem to typically impact on young people in much the same way irre‐ spective of their upper secondary graduation background. But despite the more mixed picture mediated by the results obtained for the 1993 cohort, there are also distinct country‐specific features com‐ mon to the results produced for the 1993 and 1998 cohorts, implying that much the same overall pattern can be identified for both cohorts. In par‐ ticular, early and late completers tend to fare equally well in Denmark, while late completers fare worse than young completers in Finland and Norway, and much worse in Sweden. Moreover, the age of late completion does not seem to matter that much: the “activity” (education + employ‐ ment) share is approximately the same and, hence, also the NEET share. However, there is considerable variation in the composition of late com‐ pleters’ NEET activities depending on the age of graduation. The worst outcome is obtained for adult non‐completers, except for Sweden in the sense that the labour market outcome of adult non‐ completers is highly similar to that of late completers. Moreover, the distribution across main activities of adult non‐completers is, in all four countries, approximately the same in the 1993 cohort as in the 1998 cohort. This indicates that the labour market situation of adult non‐ completers looks more or less the same irrespective of cohort and changing economic environments.

222


7. Labour market outcomes in view of background In this chapter, we look into the background of our young people, with a view of trying to identify factors that seem to be especially strongly re‐ lated to the labour market outcomes observed at three points in time: when these young persons turned 21, 26 and 31. Our statistical analysis is based on the use of so‐called multinomial logit models which show the probability of belonging to one out of several mutually exclusive groups, given a particular set of background characteristics. In our case, these groups are made up of the five categories of main activities used in the previous chapters: full‐time student, employed, unemployed, disability beneficiary or outside all of these activities (“other”). The background factors accounted for in these multinomial logit models divide basically into two main groups: one reflecting family background and the other early school‐to‐work‐transition patterns, that is, trajectories followed straight after completion of compulsory education up to age 20. Addi‐ tionally, we account for gender as well as cohort. Accounting also for cohort is relevant as we base our analysis on the pooled information available for all three youth cohorts under scrutiny, i.e., those young people who became 16 in 1993, 1998 and 2003, respectively. In the following, we report our results by background factor, starting with a brief notion on the role of gender and cohort. Thereafter we turn to family background and, finally, to early school‐to‐work transitions. Each sub‐chapter will also, when relevant, provide a short description of the background factor in question.

7.1 Gender and labour market outcomes We start by comparing the average outcome for young men and women. More precisely, we compare the overall probability of young women to that of young men of being a full‐time student, employed, unemployed, on disability benefits and in inactivity (“other”) when aged 21, 26 and 31. This information is gathered into Table 7.1 for all four Nordic countries under study. It is worth noting that the reported differences in the proba‐

bility of young women, when compared to young men, of showing up in either one of the five main activity categories under scrutiny reflects the situation after taking into account differences in family background and early school‐to‐work‐transition patterns, in addition to cohort. Table 7.1: Young women’s probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in average probabilities when compared to young men Labour market outcome Full‐time student age 21 age 26 age 31 Employed age 21 age 26 age 31 Unemployed age 21 age 26 age 31 Disability beneficiary age 21 age 26 age 31 Other (inactive) age 21 age 26 age 31

Denmark

Finland

Norway

Sweden

0.079 0.048 0.010 ‐0.084 ‐0.053 ‐0.011 0 0.006 0.005 0 0 0 0.008 0 0

0.131 0.013 0.010 ‐0.124 ‐0.053 ‐0.054 ‐0.029 ‐0.020 ‐0.015 ‐0.001 ‐0.003 ‐0.003 0.023 0.063 0.062

0.106 0.021 0.027 ‐0.089 ‐0.012 ‐0.022 ‐0.028 ‐0.014 0 0.003 0.005 0.006 0.007 0 ‐0.009

0.124 0.065 0.048 ‐0.107 ‐0.055 ‐0.050 ‐0.021 0 0.004 0 0.005 0.013 0.004 ‐0.011 ‐0.014

Notes: The reported average probabilities are obtained after also accounting for cohort, family background and early school‐to‐work‐transition patterns. A negative sign implies a weaker probabil‐ ity of young women than of young men of showing up in the labour market situation in question. Low (high) absolute values indicate a small (large) difference in this respect to young men. All re‐ ported probabilities are significant at the 99% level or more. Statistically insignificant and only weakly significant probabilities are set at zero. The probabilities are calculated from pooled infor‐ mation on the three youth cohorts under scrutiny. For definitions of the five main activity catego‐ ries, see Chapter 2.

Table 7.1 indicates the following. In all four countries, young women have a significantly higher probability of being in full‐time education, when compared to their male peers. All four countries also share the feature of a notable narrowing in this gender gap when moving towards age 31, with most of this change typically occurring before age 26, ex‐ cept in Denmark. The highest probability gap shows up for Finland with 21‐year‐old women having a 13% higher probability than their male peers of being enrolled in full‐time education. This probability is only slightly lower (12.4%) for Sweden, but comparatively low (about 8%) for Denmark. However, while this full‐time‐study probability in favour of young women remains comparatively high in Sweden up to age 31, it narrows remarkably in Finland and, in effect, down to the same low level (1%) as in Denmark. Conversely, young men have throughout, at all three age points, a clearly higher probability of being in working life. However, again we see a declining trend with age in the probability gap between the two gen‐

224


ders: a smoothly narrowing gap in Denmark, but a more step‐wise de‐ cline in the other three countries in the sense that most of also this nar‐ rowing seems to take place before the age of 26. In view of the high probability gap obtained for full‐time studying 21‐year‐old Finns, it is hardly surprising that Finland comes out with the highest gender gap also in the probability of 21‐year‐olds being in employment (12.4% in favour of men), again closely followed by Sweden (close to 11% in fa‐ vour of men). In Denmark and Norway, the corresponding gap is just below 9%. However, this higher probability of young men being in work‐ ing life has, in all countries, shrunk to about 5% already by age 26, in Norway to only about 1%. The difference between young men and young women in the average probability of showing up in (registered) unemployment is small or non‐ existent at all three age points investigated. Moreover, this holds true for all four countries. Even the largest differences fall below 3% (to the fa‐ vour of women) and are observed only at the age of 21 and only for three out of the four countries under study. Also the differences in average gender probabilities of being in disa‐ bility arrangements are more or less negligible in all four countries. In Denmark, there exist principally no differences between young men’s and young women’s probabilities of being on disability benefits. This pattern is repeated at all three age points. In Finland, young women have a marginally lower probability in this respect, and the difference to young men remains approximately unchanged up to age 31. In Norway, the situation is reversed. Additionally, young women tend to face a weakly increasing probability, when compared to young men, of ending up as disability beneficiaries. This pattern and trend is even more pro‐ nounced for young women in Sweden. When it comes to the probability of young men and young women moving into other types of inactivity, the differences in probabilities are, again, small or negligible. The only exception is Finland where young women have a comparatively high probability of withdrawing from both education and the labour force, when compared to their male peers. Moreover, this difference in inactivity probabilities is much higher at age 26, and also at age 31, than at age 21. Also the results for Sweden are worth commenting on in the sense that while young women tend to face an increasingly higher probability of going into disability arrangements when approaching age 31, young men tend instead to increasingly with‐ draw into other types of inactivity.


225

7.1.1

Main findings

The largest differences in young men’s and young women’s average prob‐ abilities of showing up in alternative labour market situations when aged 21, 26 and 31, respectively, relate to full‐time education and employment. In particular, young women face a much higher probability of being en‐ rolled in full‐time education, whereas young men are much more likely to enter working life. While these differences in education and employment probabilities prevail across all three age points investigated, they diminish substantially with age especially up to age 26, with more minor changes occurring beyond age 26, up to age 31. This overall pattern is discernible in all four countries, implying that there is a notable convergence over time not only across genders but also across countries. When it comes to NEET activities, the differences in gender probabili‐ ties are notably smaller. Also this pattern emerges for all four countries. For Denmark, the differences in gender probabilities are consistently minor or non‐existent, suggesting that young men and young women in Denmark are on average equally likely to become unemployed, to go on disability benefits and to withdraw into other types of inactivity. This holds, by and large, true also for Norway, except when it comes to un‐ employment. In this respect, young men face a clearly higher risk than young women, although the difference is small and vanishes with age. Finland stands out with a persistently, albeit only marginally, higher probability of young men becoming unemployed or of moving into disa‐ bility arrangements, whereas young women have a strikingly higher probability of withdrawing into other types of inactivity, especially be‐ yond age 21. A reversed pattern appears for Sweden with young women being on average more likely to go on disability benefits and young men into other (unknown) types of inactivity. While these country‐specific patterns were pointed out already in the previous chapter, they thus seem to be retained also after control for differences in cohort, family background and early school and labour market experiences.

226


7.2 Cohort and labour market outcomes Next, we look briefly at differences in the average probability for the three youth cohorts under study of being in these same five alternative labour market situations at age 21 and 26 after also controlling for gender, family background and early school‐to‐work‐transition patterns. At age 21, we can make comparisons across all three cohorts as we have observations for all of them five years after completion of compulsory school. At age 26, this comparison boils down to two cohorts (the 1993 and 1998 cohorts of 16‐year‐olds) for which we have observations up to ten years after com‐ pletion of compulsory education. For age 31, no corresponding compari‐ sons can be made as the information at this particular age concerns the oldest cohort only, i.e. the 1993 cohort of 16‐year‐olds. Again, the differences in average probabilities reported here (in Ta‐ bles 7.2 and 7.3) are those obtained after accounting for the full set of background factors. However, in contrast to the situation for gender, we see considerable variation in the relation between cohort and labour market outcomes depending on whether or not we leave out infor‐ mation on either family background or early school‐to‐work transitions. This observation points to a changing role of these two background fac‐ tors across cohorts. This lends further support to including both sets of background factors when analysing data containing information on more than one cohort and when, furthermore, trying to explore whether there are distinct differences between cohorts in labour market outcome probabilities in adulthood.

7.2.1

Differences in average labour market outcome probabilities at age 21

From Table 7.2 it is evident that the average probability of being in either one of the five alternative labour market situations when aged 21 reveals remarkable variation across cohorts within countries, as well as within cohorts across countries. The probability of being a full‐time student at age 21 has increased across cohorts in Denmark, stayed unchanged in Norway, and followed a concave‐type trend in Finland and Sweden. The Danish result is well in line with previous observations of more young Danes continuing in education and increasingly also delaying their com‐ pletion of an upper secondary degree (cf. Chapter 2). The Finnish and Swedish results, in turn, are likely to reflect the growing tendency across cohorts of young people delaying their start in higher education.


227

Table 7.2: 1998 and 2003 cohort probabilities in terms of labour market outcomes at age 21, by country; differences in average probabilities when compared to the 1993 cohort Labour market outcome

Denmark

Finland

Norway

Sweden

0.030 0.066

0.019 ‐0.019

0 0

0.020 ‐0.037

Employed cohort 1998 cohort 2003

‐0.040 ‐0.053

0 0.059

0 0.011

‐0.012 0.041

Unemployed cohort 1998 cohort 2003

0.009 ‐0.033

‐0.019 ‐0.036

0.034 0.013

‐0.028 ‐0.051

Disability beneficiary cohort 1998 cohort 2003

‐0.002 ‐0.002

‐0.002 0

0 0

0.003 0.012

0 0.022

0 ‐0.005

‐0.038 ‐0.021

0.016 0.034

Full‐time student cohort 1998 cohort 2003

Other (inactive) cohort 1998 cohort 2003

Notes: The reported average probabilities are obtained after also accounting for gender, family background and early school‐to‐work‐transition patterns. A negative sign implies a weaker probabil‐ ity of the cohort’s young people of showing up in the labour market situation in question compared to young people from the 1993 cohort. Low (high) absolute values indicate a small (large) difference in this respect to the 1993 cohort. All reported probabilities are significant at the 99% level or more. Insignificant and only weakly significant probabilities are set to zero. The probabilities are calculated from pooled information on the three youth cohorts under study. For definitions of the five main activity groups, see Chapter 2.

The average probability of 21‐year‐olds of being in working life is more or less the same for cohorts 1993 and 1998, but significantly higher for the 2003 cohort, notably in Finland and Sweden. For Denmark, on the other hand, we see a steady decline across cohorts in the probability of 21‐year‐ olds being in employment. These findings are the logical counterpart to the country‐specific trends in studying probabilities discussed above. In all countries, except Norway, the average probability of being reg‐ istered as an unemployed jobseeker when aged 21 is clearly lower in the youngest (2003) cohort than in the oldest (1993) cohort. For Finland and Sweden, this holds true also for the 1998 cohort. These different findings for the four Nordic countries, despite the fact that all of them experienced a deep economic recession in the early 1990s, may seem rather surprising. However, the interpretation of these differences in average unemployment probabilities across the three cohorts is not necessarily straightforward, for several reasons. First, unemployment refers to registered unemployment and, as pointed out earlier, young people not eligible for unemployment benefits might choose not to sign on. This behaviour may well have become more common with tightened conditions, especially for young people, for receiving unemployment benefits. Second, youth unemployment policies have increased both in

228


volume and in intensity over the past decade or so, but in quite different ways in the four Nordic countries. Also when it comes to disability benefits and other types of inactivity, the cross‐country picture appears quite mixed. In all countries, except Sweden, we see minor cross‐cohort differences in young people’s prob‐ ability of being on disability benefits already at age 21. For Sweden, we observe instead a weakly increasing probability across cohorts of 21‐ year‐olds showing up in disability arrangements. A similar trend is, in fact, observable for Sweden also when it comes to other types of inactivi‐ ty. Also for Denmark, the average probability of withdrawal from both education and the labour force when aged 21 is slightly higher in the 2003 cohort than in the older cohorts, whereas the opposite trend is discernible for Finland and Norway.

7.2.2

Differences in average labour market outcome probabilities at age 26

In the next table, we expand this picture for the 21‐year‐olds with corre‐ sponding results five years later, at age 26. As noted above, the compari‐ son across cohorts at this age is, due to data limitations, restricted to only two cohorts (cohorts 1993 and 1998). In order to facilitate our comparison of probabilities at age 21 and age 26, Table 7.3 reproduces the probabilities of the 1998 cohort at age 21, as reported in Table 7.2. This allows us to easily compare the situation of the 1993 cohort to that of the 1998 cohort at two specific age points. Table 7.3: 1998 cohort probabilities in terms of labour market outcomes at age 21 and 26, respec‐ tively, by country; differences in average probabilities when compared to the 1993 cohort Labour market outcome

Denmark

Finland

Norway

Sweden

Full‐time student cohort 1998, age 21 cohort 1998, age 26

0.030 0.028

0.019 0

0 ‐0.037

0.020 ‐0.015

Employed cohort 1998, age 21 cohort 1998, age 26

‐0.040 0.008

0 0.031

0 0.015

‐0.012 0.021

Unemployed cohort 1998, age 21 cohort 1998, age 26

0.009 ‐0.051

‐0.019 ‐0.020

0.034 ‐0.013

‐0.028 ‐0.028

Disability beneficiary cohort 1998, age 21 cohort 1998, age 26

‐0.002 0

‐0.002 0

0 0.003

0.003 0.006

Other (inactive) cohort 1998, age 21 cohort 1998, age 26

0 0.015

0 ‐0.009

‐0.038 0.031

0.016 0.016

Notes: See Table 7.2 above.


229

The overall impression mediated by Table 7.3 is that there is no clear‐ cut cross‐country pattern discernible in the labour market outcome probabilities of the two cohorts when moving from age 21 to age 26. In Norway and Sweden, 26‐year‐olds from the 1998 cohort are less likely to be in education than were 26‐year‐olds from the 1993 cohort. This could be argued to be a consequence of the deep economic recession in the early 1990s spurring young people to stay longer in education due to sluggish employment opportunities. However, this same pattern does not emerge for Finland, where the young people from the two cohorts are instead equally likely to be full‐time students still at age 26, nor for Denmark, where the higher probability of the 1998 cohort of being en‐ gaged in education is of the same magnitude at age 26 as five years ear‐ lier, at age 21. The different economic situations faced by young people from the two cohorts upon labour market entrance seem to play a more distinct role when it comes to employment probabilities. As shown in Table 7.3, the average employment probability at age 26 is in all four countries clearly higher for the 1998 cohort than for the 1993 cohort, more so in Finland and Sweden than in Denmark and Norway. Concomitantly, the 1998 cohort’s unemployment risk at this particular age is in all four countries lower than for the 1993 cohort, especially in Denmark. In Fin‐ land and Sweden, the difference in unemployment probabilities between the two cohorts is, in effect, the same at age 26 as it was at age 21. Concerning disability benefits, the differences in probabilities be‐ tween the two cohorts are minor. Yet, the trend is increasing rather than decreasing. In particular, in both Denmark and Finland, the 1998 cohort started out, at age 21, from a lower probability level than cohort 1993, but this gap was closed by age 26. In the other two countries, the proba‐ bility of going on disability benefits already at age 21 was the same (Norway) or marginally higher (Sweden), when compared to the 1993 cohort. By age 26, this risk had increased more rapidly in the 1998 co‐ hort, albeit it was still only marginally higher than for the 1993 cohort. Finally, with respect to other types of inactivity cohort 1998 comes out with a notably higher risk of being outside both education and the labour force at age 26. In Norway, there is a remarkable change between age 21 (an almost 4% higher risk for cohort 1993) and age 26 (a 3% lower risk for cohort 1993). In Sweden, this higher risk for cohort 1998 was observable already at age 21. Only in Finland is this inactivity risk lower at age 26 in the 1998 cohort than in the 1993 cohort, but only marginally so.

230


7.2.3

Main findings

The labour market outcome probabilities of young people belonging to different cohorts reveal considerable variation both within and across countries, showing no clear‐cut overall patterns whatsoever. A first ob‐ servation is that the differences in probabilities between cohorts are throughout quite small in size, implying that the probability of young peo‐ ple ending up in different labour market situations has not changed that much across cohorts after also controlling for cohort‐specific differences in family background and early school‐to‐work transitions. A second ob‐ servation is that there is typically no clear pattern of changing (or un‐ changing) probabilities across cohorts when going from age 21 to age 26. A comparison across the three cohorts of the labour market outcome probabilities of young people when aged 21 implies that the probability of being a full‐time student was lower for the 2003 cohort than for the 1998 and 1993 cohorts in both Finland and Sweden, obviously due to break years having become increasingly common before continuing in higher education. This contention is supported by working being much more likely among 21‐year‐olds in the 2003 cohort than in the two older cohorts. For Denmark, the findings are rather the opposite with increas‐ ing probabilities across cohorts of being engaged in education and de‐ creasing probabilities of being in employment when aged 21. For Nor‐ way, on the other hand, there are minor, if any, differences in cohort probabilities when it comes to studying and working. In relation to various NEET activities among 21‐year‐olds, the mix of cross‐cohort patterns within and between countries is much larger than for education and employment probabilities, which is only to be ex‐ pected in view of the results presented in previous chapters. The mostly very small or non‐existent differences in NEET probabilities across co‐ horts confirm, in turn, the impression of strikingly stable prevalence of NEET activities among young people. In particular, the risk of being out‐ side both education and the labour market – in disability arrangements or other types of inactivity – when aged 21 reveals an increasing rather than decreasing trend across cohorts. This holds true especially for Sweden, but also for Denmark. Only with respect to unemployment do we see a clearly lower risk among 21‐year‐olds from the 2003 cohort, though again with Norway being an exception to this pattern. A comparison of labour market outcome probabilities at age 26 – now based on only two cohorts – induces basically three main observa‐ tions. First, the probability of being employed at age 26 is in all four countries notably higher in the 1998 cohort than in the 1993 cohort. Second, the risk of being unemployed is clearly lower. Taken together,


231

these two observations are likely to mirror the impact of the deep eco‐ nomic recession in the early 1990s on young labour market entrants. Third, the differences between the two cohorts in the risk of being out‐ side both education and the labour force at age 26 are mostly small or minor and, in effect, very similar to those observed among the 21‐year‐ olds. This points to rather stable inactivity patterns across cohorts also in adulthood. Indeed, also at age 26 the risk of ending up in inactivity has been increasing rather than decreasing across cohort, a tendency which now shows up for all countries except Finland.

7.3 Family background and labour market outcomes Intergenerational transmission from parents to children has for long been an important academic as well as political issue. Special attention has thereby been paid to the parents’ educational and income levels. Indeed, as shown by e.g. Björklund et al. (2010) and Black and Devereux (2011) in their comprehensive reviews, there is a huge body of literature providing support for the contention that school success and, ultimately, labour market outcomes are closely related to family background. It is therefore of interest to include also in this context measures approxi‐ mating the family background situation of the young people belonging to the three cohorts investigated. We measure family background by use of a small set of traditional family background measures common to all four Nordic countries under study. More precisely, we ask whether there is a clear‐cut relation be‐ tween the family situation as measured by education and income and the child’s probability of being a full‐time student, employed, unem‐ ployed, disability beneficiary or inactive (“other”) at age 21, 26 and 31, respectively. The parents’ formal educational level is measured by means of three categories: basic, secondary and higher education. Also the (gross) income level of parents is split into three categories: low, middle and high wage‐income. While the information on educational level is given separately for the mother and the father, the wage‐income refers to the household‐level income, i.e. the sum of the parents’ wage‐ income. The parents’ education and income concern the year when the child turned 16, except for Finland as the Finnish data contains parental education information for the year 2010 only. While this family background information (mother’s and father’s ed‐ ucation, parental wage‐income level) is added jointly to our statistical model, we split the presentation of results, starting with the relation

232


between the mother’s educational level and the child’s labour market outcomes at later ages. Apart from family background information, the model also accounts for gender and cohort as well as early school‐to‐ work transitions.

7.3.1

The role of the mother’s educational level

Table 7.4 reports the extent to which the educational level of the mother is linked to the child’s later labour market outcomes. In particular, it gives the differences in probabilities for the child being in either one of the five alternative labour market situations at age 21, 26 and 31, respectively, depending on the educational level completed by the mother when the child turned 16. These differences in probabilities are reported for moth‐ ers with an upper secondary or a higher education. In other words, moth‐ ers with no post‐compulsory degree act as the reference group. Table 7.4: Probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, of children with a higher educated mother, by country; differences in probabilities when compared to children with a mother lacking a post‐compulsory education Mother’s highest educational level: secondary‐ or tertiary‐level degree

Labour market outcome Full‐time student at age 21 at age 26 at age 31

Denmark

Finland

Norway

Sweden

sec. ed. tert. ed.

sec. ed. tert. ed.

sec. ed. tert. ed.

sec. ed. tert. ed.

0.039 0.035 0

0.065 0.148 0.048

0.047 0.023 0

0.156 0.098 0.047

0.064 0.040 0

0.152 0.130 0.031

0.045 0.028 0

0.175 0.118 0.029


‐0.027 ‐0.023 0

‐0.063 ‐0.142 ‐0.049

‐0.035 ‐0.011 0

‐0.114 ‐0.065 ‐0.038

‐0.041 ‐0.026 0

‐0.122 ‐0.114 ‐0.026

‐0.019 ‐0.012 0.008

‐0.136 ‐0.101 ‐0.014

Unemployed at age 21 at age 26 at age 31

‐0.008 ‐0.005 0

‐0.011 ‐0.007 0

‐0.007 ‐0.006 0

‐0.033 ‐0.020 ‐0.010

‐0.011 ‐0.011 ‐0.007

‐0.021 ‐0.017 ‐0.012

‐0.018 ‐0.011 ‐0.006

‐0.038 ‐0.019 ‐0.011

0 0 0

0 0 0

0 0 0

0 0 0

0 0 0

0 0 0

‐ 0.004 ‐0.003 0

‐0.007 ‐0.006 ‐0.009

‐0.005 ‐0.007 0

0.009 0 0

‐0.005 ‐0.005 0

‐0.009 ‐0.011 0

‐0.011 0 0

‐0.010 0 0

‐0.004 0 0

0.006 0.009 0

Disability beneficiary at age 21 at age 26 at age 31 Other (inactive) at age 21 at age 26 at age 31

Notes: The three educational‐level categories correspond to ISCED 1–2, 3–4 and 5–6, respectively, with ISCED 1–2 (no post‐compulsory educational degree) being used as the category of reference. The Norwegian results also include a category for missing educational information which is not reported here, though. For other notes, see Table 7.2 above.


233

Table 7.4 presents results concerning the child’s probability of being engaged in education later in life that are well in line with the empiri‐ cal evidence reported in the international literature. In particular, there is a strong link between the child’s probability of continuing in education and the mother’s educational level. This relation strengthens at all three age points with the level of the mother’s completed degree: compared to mothers with no post‐compulsory degree, this link is no‐ tably stronger for mothers with an upper secondary degree and strongest for high‐educated mothers. However, the table also reveals that the role of the mother’s educational background weakens with the child growing older: for mothers with an upper secondary degree, the link is broken by age 31 while it has turned quite weak for mothers with a tertiary‐level degree. This overall pattern is repeated for all four countries, but with the strength of the relation varying a lot between the countries. The cross‐country situation is most similar at age 31, which again points to considerable convergence with age in the educa‐ tional behaviour of Nordic youth. Not surprisingly, the counterpart to this educational pattern is employ‐ ment probabilities of the child that are negatively related to the mother’s educational background. In other words, for the probability of the child being in employment – instead of continuing in education – we observe a reversed pattern: the child’s probability of having entered working life al‐ ready by age 21 declines notably with the educational level of the mother, implying that it is highest for children with a mother lacking a post‐ compulsory degree. Again, this same pattern is repeated for each of the three age points investigated, but with the link to the mother’s education weakening also in the case of employment probabilities when the child grows older. And again, the same pattern emerges for all four countries. The child’s risk of ending up in NEET activities – unemployment or inactivity, including disability benefits – is, on average, much more weakly related to the mother’s educational background than is the child’s probability of studying or working. There is, in all four countries, a negative but quite weak link between the child’s risk of experiencing unemployment and the mother’s educational background, with this rela‐ tionship weakening further when the child grows older. When it comes to the child’s probability of becoming a disability beneficiary, the moth‐ er’s educational background plays no significant role, except in Sweden where the link seems to exist but in an extremely weak mode. Also in relation to the child’s risk of ending up in other types of inactivity, the link to the mother’s educational background is minor or non‐existent. Again, the same overall pattern shows up in all four countries.

234


7.3.2

The role of the father’s educational level

Next, we turn to the corresponding results for the father’s highest com‐ pleted education. The relevant probabilities are presented in Table 7.5. By and large, the picture looks much the same, suggesting that the child’s labour market outcomes later in life are related to the mother’s and the father’s educational background in a highly similar way. Moreover, this holds true for both the sign and the strength of these relations. For all four countries, we observe studying probabilities of the child that are positive‐ ly related and working probabilities that are negatively related to the fa‐ ther’s educational background. This link strengthens markedly with the educational degree completed by the father. Moreover, while this pattern is repeated when the child turns 21, 26 and 31, also the relation to the father’s educational background tends to weaken considerably with the child growing older. Likewise, the risk of the child of ending up in NEET activities is only weakly, if at all, related to the father’s education. Table 7.5: Probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, of children with a higher educated father, by country; differences in probabilities when compared to children with a father lacking a post‐compulsory education

Father’s highest educational level: secondary‐ or tertiary‐level degree

Denmark

Labour market outcome

Finland

Norway

Sweden

sec. ed.

tert. ed.

sec. ed.

tert. ed.

sec. ed.

tert. ed.

sec. ed.

tert. ed.

0.032 0.018 0

0.098 0.136 0.031

0.035 0.019 0

0.156 0.103 0.037

0.059 0.037 0.016

0.163 0.129 0.034

0.033 0.022 0

0.180 0.100 0.027


‐0.025 0 0

‐0.101 ‐0.140 ‐0.034

‐0.028 ‐0.016 0

‐0.116 ‐0.087 ‐0.027

‐0.044 ‐0.024 0

‐0.141 ‐0.122 ‐0.035

‐0.023 ‐0.011 0.009

‐0.162 ‐0.098 ‐0.022


‐0.005 ‐0.007 0

‐0.011 0 0

0 0 0

‐0.034 ‐0.015 ‐0.010

‐0.009 ‐0.006 0

‐0.014 ‐0.010 0

‐0.005 ‐0.006 ‐0.005

‐0.022 ‐0.009 0

Disability beneficiary at age 21 at age 26 at age 31

0 0 0

0 0 0

0 0 0

0 0 0

0 0 0

0 0 0

‐0.002 ‐0.003 ‐0.004

‐0.003 0 0

Other (inactive) at age 21 at age 26 at age 31

0 0 0

0.012 0 0

‐0.005 0 0

‐0.008 0 0

0 0 0

‐0.008 0 0

‐0.004 0 0

0.007 0.009 0

Full‐time student at age 21 at age 26 at age 31

Notes: See Table 7.4 above.


235

7.3.3

The role of the parents’ wage‐income level

The third and final family‐related background factor concerns the par‐ ents’ total wage income. The probabilities of young people of showing up in alternative labour market activities at age 21, 26 and 31, respectively, given observed differences in parents’ wage‐income levels are displayed in Table 7.6. Other background factors accounted for in this context are, in line with the overall setting used, mother’s and father’s educational levels, as well as the young person’s gender, cohort and early school‐to‐ work‐transition experiences. Table 7.6: Probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, of children with higher wage‐income parents, by country; differences in probabilities when com‐ pared to children with parents located in the lowest tertile (one‐third) of the wage‐income scale Parents’ wage‐income level: 2nd tertile or 3rd tertile Denmark

Finland

Norway

2nd 3rd tertile tertile





0.016 0 0

0.067 0.038 0

0.027 0 0

0.074 0.046 0.014

0.021 0 0

0.058 0.027 0

‐0.011 0.035 ‐0.011 0.012 ‐0.010 ‐0.013


0.009 0.024 0.023

‐0.030 0 0.029

0 0.030 0.026

0 0 0.034

0.011 0.026 0.033

‐0.013 0 0.023

‐0.009 ‐0.013 0

‐0.021 ‐0.019 ‐0.009

‐0.021 ‐0.017 ‐0.016

‐0.047 ‐0.035 ‐0.030

‐0.012 ‐0.012 ‐0.008

‐0.023 ‐0.013 0

‐0.038 ‐0.066 ‐0.027 ‐0.039 ‐0.019 ‐0.026

0 0 0

0 0 ‐0.007

0 ‐0.002 ‐0.004

0 ‐0.004 ‐0.006

‐0.004 ‐0.005 0

‐0.005 ‐0.006 0

‐0.006 ‐0.012 ‐0.009 ‐0.016 ‐0.016 ‐0.023

‐0.016 ‐0.012 ‐0.009

‐0.016 ‐0.012 0

‐0.015 ‐0.015 ‐0.007

‐0.018 ‐0.019 ‐0.011

‐0.016 ‐0.013 ‐0.019

‐0.017 ‐0.009 0

‐0.018 ‐0.018 ‐0.020 ‐0.021 ‐0.018 0.018


Unemployed at age 21 at age 26 at age 31 Disability beneficiary at age 21 at age 26 at age 31 Other (inactive) at age 21 at age 26 at age 31

Sweden

0.074 0.067 0.063

0.060 0.063 0.080

Notes: The three wage‐income categories of parents refer, respectively, to the lowest, middle and highest one‐third of the wage‐income scale. For other notes, see Table 7.2 above.

A first observation based on Table 7.6 is that parents’ wage‐income level is to a varying degree related to the child’s probability of later in life being either studying, working, in unemployment, on disability benefits or in other types of inactivity, even after accounting for the parents’ edu‐ cational background. This means that the wage‐income level of parents plays an independent role for the child’s later labour market outcomes; it does not merely act as a proxy for parents’ educational background.

236


And vice versa, parents’ educational background is not necessarily a sufficient proxy for their labour market income status. Hence, these two family background factors are correlated, but no perfect correlates. Moreover, these two family background indicators are, in this context, occasionally oppositely related to the child’s later labour market out‐ comes, as discussed below. Children with higher wage‐income parents are, on average, more likely to be enrolled in full‐time education when aged 21, when com‐ pared to children with low wage‐income parents. Moreover, this positive relation strengthens with the wage‐income level of parents and, hence, is strongest for parents belonging to the highest one‐third of the wage‐ income scale. However, this relation between the child’s later probability of being engaged in full‐time studies and the parents’ wage‐income level weakens rapidly when the child grows older: for parents located in the middle part of the wage‐income scale it has disappeared before the child turns 26 and for parents in the highest one‐third of the wage‐income scale it has turned very small or insignificant by the time the child turns 31. Only for Sweden does this pattern look different. In particular, chil‐ dren of parents located in the middle part of the wage‐income scale show up with the lowest probability of being engaged in full‐time stud‐ ies at all three age points investigated. For high‐income parents, in turn, the link to the child’s probability of studying on a full‐time basis later in life has, by age 31, turned from positive to negative. Hence, the probabil‐ ity of studying when aged 31 tends to be highest for children with low‐ income parents. On the whole, though, this diverging pattern observed for Sweden builds on very small differences in the child’s studying prob‐ ability across the three parental wage‐income levels. The probability of the child of being in employment when aged 21, 26 and 31 is only weakly related to the parents’ wage‐income level. Moreo‐ ver, this relation is typically positive and increasing with the child’s age (i.e. stronger at age 31 than at age 21), which is opposite to the relation observed for the mother’s and the father’s educational background. These different types of relations are logical, though: while the parents’ educational background is likely to mirror the child’s probability of sub‐ stituting work with studies, their wage‐income level rather tells about the child’s later employment prospects. Also with respect to the child’s later employment probability we observe the same overall pattern for Denmark, Finland and Norway. Again, the pattern is different for Swe‐ den: a relatively strong, albeit still positive, relation between the child’s probability of being employed later in life and the parents’ wage‐income level. However, this difference appears mainly with respect to low‐


237

income parents, whereas the difference in the children’s later employ‐ ment prospects is minor or negligible when comparing middle‐income parents to high‐income parents. Children with low‐income parents tend to have the highest risk of experiencing unemployment later in life. This risk declines with the par‐ ents’ wage‐income level and, hence, is lowest for children with high‐ income parents. However, the link between the child’s likelihood of ex‐ periencing unemployment in adulthood and the parents’ wage‐income level turns increasingly weaker when the child grows older. It is worth noting, though, that this negative relation remains quite strong in both Finland and Sweden: children with middle‐income parents have still by age 31 an almost two per cent and children with high‐income parents an almost three per cent lower probability of being unemployed, when compared to children with low‐income parents. The child’s probability of moving later in life into disability arrange‐ ments is, at most, weakly negatively related to the parents’ wage‐income level. Moreover, when such a link exists, it typically strengthens slightly when the child grows older. In other words, the probability of the child of being on disability benefits is, on average, more strongly linked to the parents’ wage‐income when the child is 31 than when it is 21. This pat‐ tern is most pronounced for Sweden with the difference in children’s probability of being on disability benefits having by age 31 increased to 1.6% to the favour of those with middle‐income parents and to 2.3% to the favour of those with high‐income parents, when compared to the situation at age 31 of children with low‐income parents. For other types of inactivity, the overall pattern is again quite differ‐ ent, albeit still highly similar across the four Nordic countries under study. In particular, we observe a weak but significantly negative rela‐ tion between the child’s risk of ending up outside both education and the labour market and the parents’ wage‐income level. Moreover, this dif‐ ference in risks occurs between children with low‐income and children with higher‐income parents, whereas there is basically no difference in this respect between children with middle‐income and high‐income parents. Furthermore, this clearly higher risk of children with low‐ income parents of moving into inactivity remains practically unchanged up to age 31. In those few cases where this difference in risks shrinks and, ultimately, turns to zero, the underlying reason is not a true smoothing of risks across children with parents’ differing in their wage‐ income levels but rather a statistical property following from too few children with high‐income parents showing up in this particular situa‐ tion later in life.

238


7.3.4

Main findings

These results on the link between the child’s probability of showing up in alternative labour market situations later in life and the parents’ edu‐ cational and income background show, first and foremost, that each parental dimension – the mother’s and the father’s educational back‐ ground and the parents’ wage‐income level – mostly retains an inde‐ pendent and significant relation to the child’s later labour market out‐ comes also after accounting for other background factors, notably the child’s early school‐to‐work‐transition experiences. Indeed, an inde‐ pendent relation for each of the three parental dimensions shows up, in most cases, even though they are accounted for jointly, that is, with all three indicators included at the same time. Moreover, the parents’ edu‐ cational background and their wage‐income level tell slightly different stories, implying that accounting for both of them produces a fuller pic‐ ture of the link between the child’s family background and later labour market outcomes. The country‐specific results in relation to the child’s probability of continuing in education when aged 21, 26 and 31 indicate the following. The probability of the child studying when aged 21 increases with the parents’ educational level and also with their wage‐income level. The same pattern appears five years later, at age 26, and still ten years later, at age 31, but with the link turning increasingly weaker. A similar pattern shows up also when it comes to the link between the child’s later employment probability and the parents’ educational and income status. In particular, a higher educational level of the parents is linked to a lower probability of the child of moving into working life rather than continuing in education, whereas a higher wage‐income level of the parents tends to improve the employment prospects of the child in adulthood. The link between the child’s risk of ending up in NEET activities later in life and the parents’ educational and income background is, on aver‐ age, notably weaker and also much more dispersed, when compared to the corresponding link for the child’s educational and employment activ‐ ities. The link to the parents’ educational and income background for the child’s risk of experiencing unemployment when aged 21 resembles that observed for the child’s educational and employment probabilities: the link is negative, strengthens with the parents’ education and income, but weakens when the child grows older. Moreover, the child’s unemploy‐ ment risk seems to be more strongly linked to the parents’ income level than to their educational level, whereas the opposite holds typically true for the child’s educational and employment probabilities. In other


239

words, the child’s risk of experiencing unemployment in adulthood is strongest for children with low‐income parents, although this relation tends to weaken with age. Also the child’s risk of moving into disability arrangements or other types of inactivity when aged 21, or beyond, seems to be more strongly linked to the parents’ wage‐income level than to their educational back‐ ground. In the case of disability benefits, the negative link (i.e. lower probability) tends to strengthen with the parents’ educational and in‐ come levels: the risk is highest for children with low‐education/low‐ income parents and lowest for children with high‐education/high‐ income parents. For other types of inactivity, this trend of a steadily strengthening link depending on the child’s family background is mostly missing in the sense that the main difference occurs between low‐ education/low‐income parents, on the one hand, and higher‐ education/higher‐income parents, on the other hand. Another conspicu‐ ous difference between these two inactivity categories is that the child– parent link strengthens for disability benefits but weakens for other (unknown) types of inactivity when the child approaches the age of 31. Finally, it is worth noting that our results indicate that there is no ma‐ jor difference between the role of the mother’s and the father’s educa‐ tional background: the overall pattern is basically the same when it comes to both the sign and the strength of the investigated child–parent links. Another noteworthy feature is that there are more clear‐cut differ‐ ences in the strength of these links in relation to parents’ educational background than in relation to their wage‐income levels. More precisely, there are distinct differences in the child’s labour market outcome prob‐ abilities later in life when comparing children with low‐educated par‐ ents to children with higher educated parents, as well as when compar‐ ing children with secondary‐level educated parents to children with tertiary‐level educated parents. In relation to parents’ wage‐income levels, in contrast, the main and often only difference occurs between children with low‐income parents and children with higher‐income par‐ ents. In other words, the differences between children with middle‐ income and high‐income parents are often minor or negligible.

240


7.4 Early pathways through education and labour market outcomes The other major set of background factors included in our statistical model concerns what we have labelled “stylized” school‐to‐work path‐ ways for those young people who have achieved no post‐compulsory degree by the time they turn 21. Several recent studies have explored the relationship between non‐completion of upper secondary school and later labour market outcomes [e.g. Bäckman et al. (2011) for Scandina‐ via, and Bratsberg et al. (2010) and Falch and Nyhus (2011) for Nor‐ way]. In this section, we go one step further and investigate this rela‐ tionship by means of early school‐to‐work pathways. These stylized pathways – 16 in total – are defined and discussed in Chapter 5 and will, therefore, not be subject to a detailed presentation in this context. The results reported below refer to those obtained when also accounting for gender, cohort and family background. As became evident already in the previous sub‐chapter when discussing the role of family background, these early school‐to‐work trajectories are closely related to the educa‐ tion and income levels of parents. Accordingly we have reason to keep this in mind also when exploring the pathway results. However, even after accounting for differences in family background, early school‐to‐ work‐transition experiences often prove to be closely linked to the young person’s later labour market outcomes. In other words, good or bad early post‐compulsory‐school experiences may strengthen or weak‐ en the role of family background. The extent to which different school and labour market experiences early in life tend to increase or decrease young people’s probability of showing up in alternative labour market situations is the main topic of this sub‐chapter. Since our results cover four countries, five alternative labour market outcomes, three age points (21, 26 and 31) at which these outcomes are projected, and a total of 16 stylized school‐to‐work path‐ ways, the most efficient way of reporting this multidimensional amount of results is not self‐evident. All dimensions have their own value. How‐ ever, as the focus is on the 16 stylized pathways, they are also chosen as our point of departure for the categorisation of the results to be report‐ ed next. In other words, for each stylized pathway we will show the probability of young people following that particular pathway of being in either one of the five main activity categories investigated when aged 21, 26 and 31, respectively. Evidently, it is not possible or even meaningful to comment in detail on this multitude of results. Instead, the text will mainly focus on disen‐


241

tangling overall patterns and trends, while country‐specific results and peculiarities are easily identified in the separate tables and, therefore, mostly overlooked when commenting on key results. In the section “Main findings” concluding this sub‐chapter, an attempt will be made to summarize all these results from an “opposite” perspective, viz. from the view of young peoples’ probabilities of showing up in the five alternative labour market situations given the stylized pathway followed straight after completion of compulsory school.

7.4.1

Early post‐compulsory‐school pathways: continue in education

As noted in Chapter 5, large shares of the youngsters identified as non‐ completers at age 21 have continued in education straight after complet‐ ing compulsory school and have also stayed as full‐time students over the next five years, at least. But despite this unbroken record of study years, they have not succeeded in completing an upper secondary de‐ gree by the time they turn 21. This holds true also for another stylized pathway, the dominant feature of which is a delayed start in post‐ compulsory education due to a break year spent outside both education and the labour market (in unknown activities), upon which follows an unbroken record of years enrolled in full‐time education. Next, we look in more detail into the link between these two education‐dominated pathways and the probability of young people having followed such pathways of being in the five different labour market situations when aged 21, 26 and 31, respectively. We start by reporting the results for the early post‐compulsory‐ school trajectory representing an unbroken record of study years from age 16 up to age 20 [11111; with 1 standing for full‐time student]. As shown in Chapter 5, the share of young non‐completers following this stylized pathway – or highly similar pathways – is close to 35% for Denmark, about 30% for Norway, some 29% for Sweden but less than 22% for Finland. Table 7.7a gives the differences in the probability of these particular non‐completers, when compared to completers, of being in the five alternative labour market situations at the three different age points investigated.

242


Table 7.7a: “Study‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities compared to completers Denmark

Finland

Norway

Sweden

[11111] share 34.9%

[11111] share 21.6%

[11111] share 30.3%

[11111] share 29.2%

Full‐time student age 21 age 26 age 31

0.296 ‐0.126 0

0 ‐0.028 0

‐0.044 ‐0.093 0

0 0 0.027

Employed age 21 age 26 age 31

‐0.307 0.057 ‐0.107

‐0.042 ‐0.093 ‐0.117

‐0.034 0 ‐0.067

‐0.140 ‐0.124 ‐0.139

Unemployed age 21 age 26 age 31

0 0.032 0

0 0.059 0.047

0.045 0.041 0.034

0.121 0.064 0.043

Disability beneficiary age 21 age 26 age 31

0 0.005 0.021

0.007 0.019 0.034

0.024 0.036 0.038

0.011 0.028 0.042

Other (inactive) age 21 age 26 age 31

0.012 0.032 0.068

0.055 0.044 0.036

0.010 0.019 0

0.014 0.027 0.027


Notes: The reported average probabilities are obtained after also accounting for gender, cohort and family background. A negative sign implies a weaker probability of being in that particular activity, when compared to completers, that is, those having achieved a post‐compulsory degree by the time they turn 21. Low (high) absolute values indicate a small (large) difference with respect to complet‐ ers. All reported probabilities are significant at the 99% level or more. Insignificant and only weakly significant probabilities are set at zero. The probabilities are calculated from pooled information on the three youth cohorts under study. For definitions of the five main activity groups, see Chapter 2. For a definition and discussion of the 16 stylized post‐compulsory‐school pathways up to age 20 constructed for the Nordic non‐completers, see Chapter 5. The percentage share for the pathway in question among the non‐completers, as reported for each country, is taken from Chapter 5. “1” refers to being a full‐time student.

The corresponding results for non‐completers having delayed their start in upper secondary education by one year after completion of compulso‐ ry school [51111; with 1 standing for full‐time student and 5 for inactivi‐ ty (“other”)] are shown in Table 7.7b. While the share of non‐completers following this type of pathway is notably lower compared to the “study‐ track” non‐completers, it is far from negligible: 4.1% for Denmark, 6.2% for Norway, 7.7% for Sweden and as high as 13.5% for Finland.


243

Table 7.7b “Delayed‐study‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities compared to completers

Denmark

Finland

Norway

Sweden

[51111] share 4.1%

[51111] share 13.5%

[51111] share 6.2%

[51111] share 7.7%


0.199 ‐0.108 0

‐0.027 ‐0.047 0

‐0.059 ‐0.138 0

0.038 ‐0.033 0


‐0.275 ‐0.022 ‐0.111

‐0.096 ‐0.172 ‐0.216

‐0.078 0 ‐0.139

‐0.235 ‐0.159 ‐0.180


0.022 0.052 0

0.014 0.101 0.075

0.057 0.063 0.051

0.125 0.088 0.056


0.003 0.010 0

0.020 0.052 0.075

0.039 0.060 0.069

0.053 0.065 0.085


0.051 0.069 0.059

0.089 0.066 0.079

0.040 0.031 0.032

0.019 0.038 0.026


Notes: See Table 7.7a above. “1” refers to being a full‐time student, “5” to belonging to the catego‐ ry “other” (inactivity).

The beyond‐age‐20 studying behaviour of study‐track and delayed‐ study‐track non‐completers is quite similar. In both Finland and Sweden, these two non‐completer groups appear to have an, at most, only slightly lower probability, when compared to completers, of pursuing full‐time studies also after age 20. In Norway, on the other hand, both non‐ completer groups have a much lower probability of being in full‐time education still at age 21, with this difference with respect to completers widening further up to age 26. Moreover, the delayed‐study‐track non‐ completers seem to be in a weaker position in this respect than are the study‐track non‐completers. For Denmark, there is no clear‐cut age‐ related trend discernible: non‐completers have a much higher studying probability at age 21, but a much lower studying probability at age 26, when compared to young people having graduated from upper second‐ ary school already by age 21. Moreover, while Danish study‐track non‐ completers are more likely to continue in education at age 21 than are their delayed‐study‐track non‐completer peers, the situation is reversed at age 26. Hence, this cross‐country picture looks quite messy with an over‐ whelming dominance of dissimilarities rather than similarities. Ulti‐ mately, this is only to be expected, though, in view of highly different

244


country rates of young non‐completers, late completers and adult non‐ completers (cf. Chapter 6) in combination with distinct differences in post‐compulsory education systems. By age 31, however, these differ‐ ences across the four Nordic countries in relation to study‐dominated tracks have disappeared. The most conspicuous pattern when it comes to beyond‐age‐20 em‐ ployment probabilities is that completers are much more likely to be employed at age 21 than are study‐track non‐completers, with the em‐ ployment probability typically being even lower for delayed‐study‐track non‐completers. Moreover, this pattern is, by and large, repeated at all three age points investigated. For Finland, we observe a clear‐cut trend in this respect with the employment probability of these non‐completer groups weakening considerably with age. A similar age‐related pattern is not discernible for the other three countries. The counterpart to these differences in employment probabilities is a much lower unemployment risk among completers than among both study‐track and delayed‐study‐track non‐completers, with the latter group of non‐completers again doing worse than their study‐track non‐ completer peers. However, this difference with respect to completers in unemployment risks tends to narrow with age, a pattern that is discern‐ ible in all four countries and most clearly for Sweden. The risk of becoming a disability beneficiary is notably higher for de‐ layed‐study‐track non‐completers, less so for study‐track non‐ completers, when compared to completers. Moreover, this risk is in‐ creasing with age, more strongly for delayed‐study‐track non‐ completers than for study‐track non‐completers. This overall pattern shows up in all four countries. The cross‐country pattern is more dis‐ persed when it comes to other (unknown) types of inactivity. Study‐ track and especially delayed‐study‐track non‐completers encounter a higher risk than completers also in this respect, but the age‐related trend varies a lot across the four countries: in Denmark it is increasing and in Finland decreasing with age. These differences in risks are, on average, smaller in Norway and Sweden than in Denmark and Finland, revealing no age‐related trend whatsoever.

7.4.2

Early post‐compulsory‐school pathways: move into working life at a young age

The analysis in Chapter 5 of non‐completers’ stylized post‐compulsory‐ school trajectories revealed that large shares of them drop out from upper secondary education and move into working life , so it seems, on


245

quite a permanent basis. This dropout pattern is particularly common among those having continued in upper secondary education for two or three years. It is less frequent among young non‐completers having dropped out already after one year, as well as among those who never continued in education after compulsory school. Moreover, although they tend to spend most of their time after school dropout in employ‐ ment, their employment profile is of a much more bumpy nature, when compared to that of non‐completers dropping out only after two or three years in upper secondary education. Next we investigate to what extent these employment‐dominated post‐compulsory‐school pathways of non‐completers are related to their probability of ending up in the five main labour market situations under scrutiny 5, 10 and 15 years after completion of compulsory education. Again, the group of reference is completers, that is, those young people having achieved an upper secondary degree by the time they turn 21. We thereby start with the most typical early employment tracks fol‐ lowed by non‐completers, viz. tracks characterised by the young non‐ completer continuing directly in post‐compulsory education for two or three years before substituting school with work, but without having graduated from upper secondary education [11222 and 11122; with 1 standing for full‐time student and 2 for employed]. As indicated in Chapter 5, the share of non‐completers following these two types of early employment pathways is close to 24% for Nor‐ way and only slightly lower for Denmark (22%), Sweden (21%) and Finland (about 20%). These four Nordic countries also share the feature of the 3‐year‐study–employment track covering a clearly larger share of the non‐completers than the 2‐year‐study–employment track, except in Denmark where the latter track is slightly more common. Table 7.8a presents the differences for the relevant probabilities, when compared to the situation of completers, for non‐completers leaving upper second‐ ary education for work after three years of full‐time studies. The corre‐ sponding information for non‐completers dropping out for employment already after two years of full‐time studies is given in Table 7.8b. Not surprisingly, young non‐completers moving early into working life are less likely than completers to show up as full‐time students also be‐ yond age 20. In all four countries, this lower studying probability, when compared to completers, is highest at age 21, considerably lower at age 26 and almost non‐existent at age 31. In view of the results presented in Chapter 2, this time trend obviously follows from increasing numbers of completers moving into working life rather than from early dropouts in working life returning to full‐time education. Interestingly, both the pat‐

246


tern and strength of this lower studying probability are highly similar for non‐completers staying either two or three years in upper secondary edu‐ cation before dropping out. Hence, it does not seem to matter that much whether the young non‐completer has left upper secondary education after two or only after three years of full‐time studies. Table 7.8a: “Dropout‐after‐three‐years‐employment‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabili‐ ties when compared to completers Denmark

Finland

Norway

Sweden

[11122] share 10.3%

[11122] share 11.1%

[11122] share 14.6%

[11122] share 13.8%


‐0.234 ‐0.119 0.031

‐0.294 ‐0.081 0

‐0.406 ‐0.114 ‐0.029

‐0.248 ‐0.089 0


0.149 0.033 ‐0.095

0.266 0.043 0

0.290 0 ‐0.067

0.204 0.026 ‐0.066


0.028 0.033 0

0 0.021 0

0.058 0.047 0.041

0.045 0.041 0.030


0 0.005 0.014

‐0.003 0 0

0.011 0.015 0.029

0 0.009 0.023


0.055 0.043 0.039

0.028 0 0

0.047 0.031 0.026

0 0.013 0.016


Notes: See Table 7.7a above. “1” refers to being a full‐time student, “2” to being employed.

Likewise, it is not surprising to find that non‐completers moving early into working life tend to have a higher probability of being employed also at age 21, when compared to completers. This advantage in terms of employment prospects vanishes rapidly, though. In all four countries, the employment probability of these non‐completers is at age 26 only slightly higher than among completers and by age 31, it has turned nega‐ tive in all countries except Finland. In other words, young completers tend to rapidly close the employment gap to non‐completers having moved into working life at an early age. Already by age 31, the employ‐ ment prospects of completers are clearly better. Again, we see small, if any, differences in these respects between non‐completers having left upper secondary education after two years compared to those having dropped out only after three years.


247

Table 7.8b: “Dropout‐after‐two‐years‐employment‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities when compared to completers Denmark

Finland

Norway

Sweden

[11222] share 11.8%

[11222] share 8.8%

[11222] share 9.4%

[11222] share 7.4%


‐0.205 ‐0.131 0

‐0.294 ‐0.105 0

‐0.418 ‐0.194 ‐0.043

‐0.224 ‐0.070 0


0.119 0.037 ‐0.080

0.256 0.039 0

0.293 0.078 ‐0.053

0.169 0 ‐0.061


0.030 0.040 0.019

0 0.022 0

0.063 0.051 0.030

0.055 0.036 0.026


0.002 0.008 0.013

0 0 0

0.016 0.027 0.031

0 0.015 0.024


0.053 0.047 0.039

0.030 0.042 0

0.047 0.038 0.035

0 0.013 0.020


Notes: See Table 7.7a above. “1” refers to being a full‐time student, “2” to being employed.

The probability of these early‐working non‐completers of encountering unemployment at age 21, or beyond, is only slightly higher compared to the situation faced by their completer peers. Moreover, this pattern of relatively small differences in unemployment risks, when compared to completers, holds true at all three age points investigated and, indeed, shows up in all four countries. Also in this respect, the situation is much the same for young non‐completers moving into employment‐ dominated tracks after three years or already after two years in upper secondary education. The two tables also indicate that these non‐completers have an only a marginally, if any, higher probability of going into disability arrange‐ ments after having turned 20, when compared to completers. For Fin‐ land, there are no significant differences in this respect between com‐ pleters and non‐completers having moved early into working life. In the other three countries, the difference is small, albeit weakly increasing with age. Again, the pattern is strikingly similar for those dropping out either after two or only after three years in upper secondary education. When it comes to other types of inactivity, the disadvantage in rela‐ tion to completers is clearly stronger, especially in Denmark and Nor‐ way. However, this higher risk of non‐completers having entered work‐

248


ing life at a young age of ending up in inactivity declines with age, imply‐ ing that the difference with respect to completers is typically smaller at age 31 than at age 21. On the whole, then, we identify relatively small differences in the risk of showing up in NEET activities between completers and non‐ completers moving, so it seems, quite successfully into working life already at a young age. This finding is most likely related to their comparatively good employment prospects, when compared to com‐ pleters, also later in life. Hence, by means of a close to finished upper secondary education coupled with early engagement in working life, they seem to have been able to achieve competencies that keep them employed also in adulthood. Moreover, in this respect we see no ma‐ jor differences between non‐completers dropping out after two or only after three years in upper secondary education. Next we compare the outcome of these two groups of non‐ completers to those of their non‐completer peers having dropped out already after one or no year in post‐compulsory education. More pre‐ cisely, this first post‐compulsory‐school year is typically spent either in education [12222] or outside both education and the labour mar‐ ket [52222], with 1 standing for full‐time student, 2 for employment and 5 for “other” (inactivity). The results for both of these groups are included in Table 7.8c for the simple reason that the two groups are so small for Norway that it is not possible to obtain separate results for them at age 21. Indeed, also when taken together, they cover less than 4% of the Norwegian non‐completers. Their combined share is slightly higher in Sweden (about 5%) and highest in Denmark (13.6%) and Finland (10.6%). However, while those dropping out from school already after one year represent the dominating track (close to 10%) of the two in Denmark, the “non‐starters” make up a larger share (about 6%) in Finland. As is also to be expected, the studying probability of these non‐ completers looks much the same as the studying probability of non‐ completers dropping out slightly later. All four groups of non‐completers substitute school with work at a young age, implying that they are much less likely than their completer peers to be enrolled in education espe‐ cially when comparing their situation at age 21. With increasing num‐ bers of also completers moving into working life, the difference in study‐ ing probabilities between completers and non‐completers narrows rap‐ idly and, ultimately, disappears.


249

Table 7.8c: “Drop‐out‐after‐one‐year‐employment‐track” & “Inactivity‐year‐employment‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respec‐ tively, by country; differences in probabilities when compared to completers

Denmark


[12222] [52222] Share share 9.7% 3.9%

Finland

Norway

Sweden

[12222] [52222] share share 4.4% 6.2%

[12222] [52222] share share 2.7% 0.9%

[12222] [52222] share share 3.0% 2.1%


‐0.172 ‐0.154 0

‐0.286 ‐0.174 0

‐0.288 ‐0.110 0

‐0.350 ‐0.129 0

‐0.383 ‐0.215 ‐0.065

‐0.383 ‐0.138 0

‐0.217 ‐0.102 0

‐0.290 ‐0.094 0


0.099 0.061 ‐0.067

0.171 0.044 ‐0.107

0.224 0 0

0.271 0 0

0.195 0.070 ‐0.056

0.195 0 ‐0.139

0.146 0 ‐0.080

0.236 0 ‐0.108


0.024 0.034 0.013

0.050 0.043 0

0 0.026 0

0 0.032 0

0.065 0.066 0.047

0.065 0.063 0.051

0.047 0.038 0

0 0.041 0.042

Disability beneficiary at age 21 at age 26 at age 31

0.002 0.009 0.014

0 0.010 0.021

0 0 0

0 0 0

0.024 0.021 0.028

0.024 0.060 0.069

0 0 0

0.017 0.018 0

Other (inactive) at age 21 at age 26 at age 31

0.047 0.049 0.033

0.062 0.078 0.068

0.048 0.047 0

0.058 0.054 0

0.099 0.059 0.046

0.099 0.031 0.032

0 0.050 0

0.020 0.031 0

Notes: See Table 7.7a above. “1” refers to being a full‐time student, “2” to being employed and “5” to belonging to “other” (inactivity). For Norway, the distribution across the five labour market outcomes of non‐completers following an early post‐compulsory‐school trajectory of [12222] or [52222] is too skewed at age 21 to allow separate probabilities to be estimated for the two pathways.

Also with respect to the probability of being employed still beyond age 20, the overall pattern is highly similar for the four groups of non‐ completers following early employment‐dominated tracks. More pre‐ cisely, all of them tend to have a higher probability of being employed also at age 21, when compared to completers. Another common feature of these four non‐completer groups is that this advantage in terms of employment prospects vanishes rapidly with age. In all four countries, they have when aged 26 an employment probability that is only slightly, if at all, higher than for completers. By age 31, this advantage has turned into a disadvantage in all countries except Finland. However, despite this common overall pattern there are also distinct differences between the four groups of early‐employment non‐completers, the most conspicuous being the following: young non‐completers not having continued in post‐ compulsory education have, when aged 31, notably weaker employment prospects not only when compared to completers but also when com‐ pared to their non‐completer peers having continued in post‐ compulsory education for at least one year. In Norway, for instance,

250


these non‐starters’ employment probability at age 31 is 14% lower than for completers and about 8% lower than for their dropout non‐ completer peers. When it comes to NEET activities, similar overall patterns show again up for all four non‐completer groups following early employment‐ dominated tracks. Broadly speaking, all of them face a higher risk of ending up in NEET activities, when compared to completers. However, as is evident in Table 7.8c illustrating the outcome for non‐completers entering working life after only one or no year in post‐compulsory edu‐ cation, the results with respect to NEET activities are occasionally sur‐ rounded by uncertainty. Compared to the results reported in Tables 7.8a and 7.8b, this shows up in less clear‐cut age‐related trends. More im‐ portant, it is reflected in a considerable number of cells containing zeros, indicating no difference in probabilities when compared to completers, when the correct interpretation rather is that the number of completers and/or non‐completers in that particular labour market situation is too small for producing reliable (robust) results. But also with this warning in mind, the labour market outcome probabilities reported in the three tables seem to strongly suggest that young non‐starters in upper sec‐ ondary education fare much worse as young adults also when compared to their working peers having dropped out early from upper secondary education. Moreover, this contention is well in line with the results re‐ ported in the previous chapter.

7.4.3

Early post‐compulsory‐school pathways: unemployment experiences at a young age

In our national datasets, “unemployment” covers situations where the young person is registered as an unemployed jobseeker. This means that we occasionally observe quite large differences across the four Nordic countries in the shares of young people in unemployment, but also in the cross‐cohort trend in these shares within single countries. This holds true for those below the age of 25 and even more so for those below the age of 21. Needless to say, a major reason for these observations is differences between countries and changes within countries in the preconditions for being eligible for unemployment benefits in combination with potential differences between countries and changes within countries in the regis‐ tering behaviour of particularly those young people not eligible for receiv‐ ing unemployment benefits: Will they still register as unemployed? In Chapter 5, we constructed for the non‐completers a total of three stylized post‐compulsory‐school trajectories dominated by unemploy‐


251

ment experiences at a young age. These tracks are identical to the em‐ ployment‐dominated tracks discussed above in the sense that they start with a varying number of years in post‐compulsory schooling before dropping out without graduating. However, now a dropout after one, two or three years in post‐compulsory schooling turns into early years dominated by unemployment. Next, we present and comment on the labour market outcomes up to age 31 that are most probably related to dropping out from school and experiencing prolonged unemployment already before the age of 21. Again, we start by first presenting results for the most typical track: non‐completers who continue directly in post‐compulsory education but drop out after three years of full‐time studies just to end up in (regis‐ tered) unemployment [11133; with 1 standing for full‐time student and 3 for unemployment]. Based on the information given in Chapter 5, the share of non‐completers in our three youth cohorts following this type of unemployment‐dominated pathway is 2.1% for Denmark, 3.5% for Finland, 4.5% for Norway and 11.6% for Sweden. Table 7.9a reports the differences for the relevant probabilities when compared to the situa‐ tion of completers. Table 7.9a “Dropout‐after‐three‐years‐unemployment‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities when compared to completers

Denmark

Finland

Norway

Sweden

[11133] share 2.1%

[11133] share 3.5%

[11133] share 4.5%

[11133] share 11.6%


‐0.119 ‐0.124 0

‐0.181 ‐0.068 0

‐0.295 ‐0.164 ‐0.049

‐0.129 ‐0.048 0


‐0.173 ‐0.086 ‐0.236

‐0.068 ‐0.143 ‐0.172

0.094 0 ‐0.111

‐0.147 ‐0.151 ‐0.198


0.141 0.062 0.037

0.128 0.113 0.090

0.102 0.086 0.069

0.225 0.107 0.074


0.015 0.049 0.066

0 0 0

0.028 0.041 0.041

0.011 0.040 0.059


0.135 0.098 0.081

0.118 0.083 0

0.072 0.060 0.050

0.040 0.052 0.053


Notes: See Table 7.7a above. “1” refers to being a full‐time student, “3” to being unemployed.

252


The pattern in relation to the probability of showing up as a full‐time student at age 21, 26 and/or 31 is identical to that observed for young non‐completers substituting school with work: a much lower probability of being enrolled in full‐time education at age 21, with this difference with respect to completers typically levelling out with age. Dropout non‐ completers with early unemployment experiences also tend to encoun‐ ter much lower employment possibilities. These notably weaker em‐ ployment prospects are discernible already at age 21. Moreover, they tend to weaken further with age, being considerably weaker at age 31 than at age 21. Conversely, their risk of continuing in or returning to unemployment is obvious. Indeed, they encounter still at age 31 a risk of becoming unemployed that is higher than that faced by completers, but also higher than the risk of becoming unemployed characterising non‐ completers having substituted school with work at an early age. When it comes to other types of NEET activities, we observe initially small but increasing risks of these non‐completers moving into disability arrangements. This pattern shows up in all countries except Finland. Non‐completers following unemployment tracks after having dropped out after three years in upper secondary education also have a remarka‐ bly high risk of being in other (unknown) types of inactivity already when aged 21. While this risk tends to decline with age it is, nonetheless, relatively high still at age 31. In Denmark, for instance, it is almost 12% higher than for completers and, indeed, also much higher than for non‐ completers having dropped out from school into working life. Dropping out into unemployment after three years in upper second‐ ary education without graduation thus tends to result in a risky labour market situation also as a young adult. In particular, the employment prospects are much weaker and the risk of ending up in NEET activities much higher than for completers, but also much higher than for non‐ completer peers having successfully entered the labour market despite dropping out from upper secondary education. Moreover, this weaker situation typically prevails still 15 years after completion of compulsory school. These findings thus lend further support to the international literature in this field. Next, we compare this outcome for young unemployed non‐ completers having dropped out after three years of upper secondary education to that of their unemployed peers having dropped out already after two years or after only one year in post‐compulsory education. Table 7.9b reports the results obtained for non‐completers who contin‐ ued in post‐compulsory education but dropped out after two years of full‐time studies just to end up in unemployment [11333; with 1 stand‐


253

ing for full‐time student and 3 for unemployment]. The share of non‐ completers following this type of pathway is of much the same size as for the [11133] pathway, but only in Denmark and Finland. For Norway it is slightly lower and for Sweden notably lower. Table 7.9b: “Dropout‐after‐two‐years‐unemployment‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities when compared to completers

Denmark

Finland

Norway

Sweden

[11333] share 1.8%

[11333] share 3.7%

[11333] share 3.0%

[11333] share 4.2%


‐0.202 ‐0.158 0

‐0.197 ‐0.074 0

‐0.399 ‐0.218 0

‐0.112 ‐0.060 0


‐0.157 ‐0.136 ‐0.214

‐0.087 ‐0.188 ‐0.174

0.150 0 ‐0.137

‐0.188 ‐0.190 ‐0.224


0.159 0.088 0

0.128 0.138 0.109

0.104 0.089 0.072

0.212 0.136 0.081


0.037 0.076 0.039

0 0 0.026

0.029 0.040 0.047

0.018 0.050 0.073


0.162 0.130 0.118

0.151 0.111 0.062

0.115 0.073 0.037

0.069 0.064 0.058


Notes: See Table 7.7a above. “1” refers to being a full‐time student, “3” to being unemployed.

Table 7.9c, finally, reports the probability of ending up in alternative labour market situations for non‐completers starting an unemployment track record already after one year in post‐compulsory education [13333, with 1 standing for full‐time student, 3 for unemployment]. As is evident in the table, the share of non‐completers in our data following such tracks is mostly very low: 0.7% for Denmark, 0.8% for Sweden, 1.9% for Norway but as high as 5.1% for Finland.

254


Table 7.9c: “Dropout‐after‐one‐year‐unemployment‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabili‐ ties when compared to completers Denmark

Finland

Norway

Sweden

[13333] share 0.7%

[13333] share 5.1%

[13333] share 1.9%

[13333] share 0.8%


‐0.211 ‐0.106 0

‐0.167 ‐0.105 0

‐0.349 ‐0.174 0

‐0.193 ‐0.076 0


‐0.146 ‐0.178 ‐0.181

‐0.108 ‐0.202 ‐0.255

0.091 0 ‐0.162

‐0.197 ‐0.192 ‐0.259

0.131 0.130 0

0.109 0.174 0.117

0.112 0.095 0.082

0.196 0.106 0.091

0 0 0

0.006 0.023 0.033

0.029 0.041 0.050

0.042 0.073 0.081

0.211 0.122 0.090

0.160 0.110 0.123

0.118 0.101 0.060

0.152 0.089 0.091


Unemployed age 21 age 26 age 31 Disability beneficiary age 21 age 26 age 31 Other (inactive) age 21 age 26 age 31

Notes: See Table 7.7a above. “1” refers to being a full‐time student and “3” to being unemployed.

By and large, the same overall pattern is repeated for all three groups of non‐completers with early experiences from (registered) unemploy‐ ment. This contention seems to hold true despite the uncertainty evi‐ dently surrounding the results reported in Table 7.9c due to small shares of non‐completers following such tracks (except for Finland). In sum, the picture mediated by the three tables indicates the following. The probability of these young non‐completers of being enrolled in full‐ time education when aged 21 is low, but this difference with respect to completers diminishes with graduated young people moving increasing‐ ly from education into working life. All these non‐completers going early into unemployment also face a much lower probability of being em‐ ployed by age 21, and beyond, with their employment prospects wors‐ ening further with age. A comparison of the three tables further suggests that the picture outlined above tends to strengthen when going from young unemployed non‐completers having dropped out only after three years in upper sec‐ ondary education to those having dropped out already after one year in post‐compulsory education. However, such a pattern appears only in relation to education and employment probabilities. When it comes to the risk of experiencing unemployment also in adulthood, there are mi‐


255

nor differences between these three groups of young unemployed non‐ completers differing in their years in post‐compulsory education before dropout. Indeed, this seems to hold true also with respect to the risk of moving on disability benefits, whereas the risk of ending up in other (unknown) types of inactivity reveals a tendency of typically being the higher, the earlier the young non‐completer has dropped out from school into unemployment. Finally, while the three tables indicate the difference in probabilities when compared to completers, a comparison of Tables 7.9a‐c and Tables 7.8a‐c suggests that young non‐completers with early experiences from unemployment fare, on average, worse also when compared to young non‐completers skipping school for work.

7.4.4

Early post‐compulsory‐school pathways: becoming a young disability beneficiary

When it comes to young disability beneficiaries, we refer to a recent analysis of ours (Albæk et al., 2014) published in the final report of the so‐called NORWELL project funded by the Nordic Council of Ministers. Nonetheless, we will use some space also in this context to highlight labour market outcomes of young adults having experienced health problems at an early age. Doing so, we are able to obtain a fuller picture of the role of all the 16 stylized pathways presented and discussed in Chapter 5 in terms of non‐completers’ labour market situation as young adults. Moreover, the Albæk et al. (2014) analysis did not cover Sweden, merely Denmark, Finland and Norway. As for the employment and unemployment pathways discussed above, we concentrate on early tracks starting with a varying number of years in post‐compulsory schooling before ending with the young per‐ son dropping out, now to become a disability beneficiary. The total share of non‐completers in these types of pathways is low relative to the shares of non‐completers following the other stylized pathways – close to 1% for Denmark, 3% for Finland, about 2% for Norway and 4% for Sweden – but far from negligible in absolute terms. The first stylized pathway analysed below turns into a disability‐ benefit track after three years in post‐compulsory schooling straight after completion of basic education [11144; with 1 standing for full‐time stu‐ dent and 4 for disability beneficiary]. The share of young non‐completers following such pathways is 0.1% for Denmark, 0.5% for Finland, 1.5% for Norway and 1% for Sweden. Table 7.10a gives the differences for the rele‐ vant probabilities when compared to the situation of completers.

256


Table 7.10a: “After‐three‐years‐dropout‐disability‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabili‐ ties when compared to completers Denmark

Finland

Norway

Sweden

[11144] share 0.1%

[11144] share 0.5%

[11144] share 1.5%

[11144] share 1.0%


‐0.427 ‐0.330 ‐0.097

‐0.378 ‐0.152 0

‐0.188 0 0

‐0.172 ‐0.107 0


‐0.440 ‐0.353 0

‐0.287 ‐0.526 ‐0.504

‐0.109 ‐0.160 ‐0.182

‐0.384 ‐0.580 ‐0.753


‐0.025 ‐0.033 ‐0.016

‐0.061 0 0

‐0.139 0.107 0.072

‐0.038 0 ‐0.021

0.945 0.716 0

0.702 0.599 0.501

‐0.074 0.099 0.120

0.612 0.728 0.760

‐0.052 0 ‐0.027

0 0 0

‐0.084 0 0

‐0.018 0.026 ‐0.023


Disability beneficiary age 21 age 26 age 31 Other (inactive) age 21 age 26 age 31

Notes: See Table 7.7a above. “1” refers to being a full‐time student, “4” to being a disability benefi‐ ciary. For Norway, the distribution across the five labour market outcomes of non‐completers following an early post‐compulsory‐school trajectory of [11144], [11444] or [14444] is too skewed at age 21 to allow separate probabilities to be estimated for these three pathways at this particular age, which explains the parentheses surrounding the estimated difference in probabilities for alter‐ native labour market outcomes at age 21.

Young non‐completers moving into disability arrangements after having dropped out from three years of full‐time studies at an upper secondary school reveal a low probability of having returned to school by age 21. They also have a low probability of being employed at this particular age, with their employment prospects weakening further with age. The only exception is Denmark, but due to the low share of non‐completers in this type of track, the Danish results should in this context be interpreted with caution. Generally non‐completers having followed this type of track also tend to face a slightly higher risk of being unemployed, when compared to completers, but this pattern varies a lot between countries and also within countries with age. Additionally, they typically have a high risk, also when compared to other young non‐completers, of continuing on or returning to disability benefits as young adults, whereas their likelihood of ending up in other types of inactivity is mostly very low. In the next table (Table 7.10b), we turn the focus to young disability beneficiaries having dropped out after two years or already after one


257

year in post‐compulsory schooling [11444 and 14444; with 1 standing for full‐time student and 4 for being a disability beneficiary]. The share of non‐completers following either one of these two pathways is com‐ paratively high only in two cases, both of which cover immediate or al‐ most immediate moves into disability arrangements after completion of compulsory school [the 14444 track]: 2.3% for Finland and 2.8% for Sweden, with the rest of shares staying below 0.5%. Table 7.10b: “Dropout‐after‐two‐years‐disability‐track” and “Dropout‐after‐one‐year‐disability‐ track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities when compared to completers

Denmark


Finland

[11444] [14444] share share 0.4% 0.4%

Norway

[11444] [14444] share share 0.2% 2.3%

Sweden

[11444] [14444] share share 0.3% 0.1%

[11444] [14444] share share 0.2% 2.8%

Full‐time stu‐ dent at age 21 at age 26 at age 31

‐0.439 ‐0.320 0

‐0.439 ‐0.330 0

‐0.346 ‐0.192 ‐0.132

‐0.419 ‐0.234 ‐0.118

‐0.188 0 0

‐0.188 0

‐0.293 0 0

‐0.379 ‐0.196 ‐0.058


‐0.395 0 0

‐0.407 ‐0.483 0

‐0.295 ‐0.556 ‐0.607

‐0.356 ‐0.602 ‐0.736

‐0.109 ‐0.468 0

‐0.109 0 0

‐0.450 ‐0.604 ‐0.639

‐0.459 ‐0.626 ‐0.812


‐0.025 ‐0.033 0

‐0.025 ‐0.033 0

‐0.070 0 ‐0.049

‐0.078 ‐0.053 ‐0.038

‐0.139 0.098 0

‐0.139 0 0

‐0.054 0 ‐0.021

‐0.070 ‐0.039 ‐0.016

0.911 0 0

0.923 0.871 0

0.720 0.715 0.839

0.883 0.925 0.933

‐0.074 0.105 0.118

‐0.074 0 0.118

0.831 0.652 0.660

0.942 0.891 0.908

‐0.052 ‐0.025 0

‐0.052 ‐0.025 0

0 0 ‐0.050

‐0.030 ‐0.036 ‐0.041

‐0.084 0.132 0

‐0.084 0 0

‐0.035 0 0

‐0.034 ‐0.032 ‐0.021

Disability beneficiary at age 21 at age 26 at age 31 Other (inactive) at age 21 at age 26 at age 31

Notes: See Table 7.10a above.

The overall impression arising from comparing the labour market pro‐ spects of the three groups of young non‐completers experiencing health problems at an early age is that their situation looks much the same, despite the uncertainty undeniably surrounding many of the results due to small numbers of observations fulfilling particular combinations of background characteristics. In other words, it does not seem to matter how many years the young person has spent in upper secondary educa‐ tion before dropping out without achieving a degree. In all three cases, the most likely outcome is that the young person continues in or returns to disability arrangements also in adulthood.

258


7.4.5

Early post‐compulsory‐school pathways: withdrawal at young age into inactivity

The last set of our 16 stylized pathways constructed for the non‐ completers contains a total of four early post‐compulsory‐school trajec‐ tories dominated by years spent outside both education and the labour force, that is, in activities not covered by any of the large administrative registers from which our datasets are compiled. These “other” situations may, as noted in previous chapters, imply engagement in highly different activities which may have a positive influence on the young person’s near‐future choices, but may also involve activities starting a risky pathway into adulthood. The first three pathways are characterised by the young non‐ completer having continued directly in education after completing com‐ pulsory school. Their main difference lies in the number of years en‐ rolled in full‐time education before dropping out: after three years, after two years or already after one post‐compulsory education year. Taken together, these three pathways cover 16.3% of the non‐completers in Denmark, 11.3% in Finland, 22.8% in Norway and 10.3% in Sweden. In other words, a substantial number of young people drop out from school just to withdraw from taking part in educational or in labour market activities. The fourth pathway, in contrast, involves no post‐compulsory education whatsoever: it starts with the young person withdrawing straightaway after leaving compulsory school from both education and the labour market, and continues with the person staying outside any such activities up to age 20, at least. The share of young non‐completers following such tracks is low in Norway (1.7%) and Denmark (3.5%) but worryingly high in both Finland (8%) and Sweden (7%). Next we look somewhat more in detail into the labour market out‐ comes that these withdrawal pathways seem to be most strongly related to. We thereby start with the pathway preceded by three years in post‐ compulsory education before dropout [11155; with 1 standing for full‐ time student and 5 for “other” (activities)]. The share of non‐completers following this type of early trajectory is strikingly high in all four Nordic countries under study: 6.4% for Denmark, 7.4% for Finland, 13.3% for Norway and 6.1% for Sweden. The probability of ending up in alterna‐ tive labour market situations in adulthood is, therefore, of particular interest for this comparatively large group of non‐completers. These probabilities, as compared to completers, are displayed in Table 7.11a. Table 7.11a shows that young non‐completers withdrawing into inac‐ tivity after having dropped out from upper secondary education after three years of full‐time studies, are unlikely to have returned to educa‐


259

tion by age 21. Although this difference in studying probabilities, when compared to completers, shrinks with age, this narrowing is, once again, obviously mainly due to completers having left the education system rather than these dropouts having returned to education by age 26 or by age 31. They are also much less likely to be in employment at age 21, with their employment prospects weakening rather than improving with age. Probably, this also explains their typically only slightly higher risk, when compared to completers, of being registered as unemployed jobseekers as young adults. Moreover, their risk of moving into disability arrangements increases with age, especially in Denmark and Sweden. Finally, they are very likely to continue in inactivity still at age 21, with this risk declining only slowly with age. Table 7.11a: “Dropout‐after‐three‐years‐withdrawal‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabili‐ ties when compared to completers Denmark

Finland

Norway

Sweden

[11155] share 6.4%

[11155] share 7.4%

[11155] share 13.3%

[11155] share 6.1%


‐0.164 ‐0.074 0.058

‐0.248 ‐0.049 0

‐0.373 ‐0.083 0

‐0.207 0 0


‐0.160 ‐0.132 ‐0.236

‐0.097 ‐0.182 ‐0.184

0.149 ‐0.072 ‐0.157

‐0.153 ‐0.192 ‐0.170


0.057 0.059 0.024

0.027 0.100 0.074

0.068 0.063 0.051

0.113 0.053 0


0.015 0.045 0.073

0.014 0.017 0.030

0.025 0.032 0.042

0.034 0.075 0.103


0.252 0.103 0.081

0.304 0.114 0.085

0.132 0.060 0.053

0.213 0.086 0.048


Notes: See Table 7.7a above. “1” refers to being a full‐time student, “5” to being in “other” activities.

The corresponding differences in probabilities, as compared to complet‐ ers, for those young non‐completers withdrawing into inactivity after two years in post‐compulsory education [11555; with 1 standing for full‐ time student and 5 for “other” (activities)] are reported in Table 7.11b. The share of non‐completers following this type of track is less than 2.5% in both Finland and Sweden, about twice as large for Denmark and, again, highest (6%) for Norway.

260


Table 7.11b: “Dropout‐after‐two‐years‐withdrawal‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabili‐ ties when compared to completers

Denmark

Finland

Norway

Sweden

[11555] share 4.9%

[11555] share 2.2%

[11555] share 6.0%

[11555] share 2.4%


‐0.165 ‐0.122 0.051

‐0.220 ‐0.069 0

‐0.341 ‐0.111 0

‐0.158 0 0


‐0.232 ‐0.176 ‐0.290

‐0.192 ‐0.281 ‐0.200

0.090 ‐0.098 ‐0.163

‐0.262 ‐0.270 ‐0.291


0.078 0.073 0.023

0 0.112 0

0.077 0.081 0.059

0.138 0.064 0.071


0.054 0.097 0.125

0.022 0 0

0.027 0.037 0.042

0.058 0.083 0.100


0.266 0.128 0.092

0.357 0.211 0

0.146 0.091 0.079

0.224 0.132 0.074



Based on Table 7.11b, the following may be concluded. Also young non‐ completers withdrawing into inactivity already after two years in upper secondary education face a much lower probability of being enrolled in education at age 21, when compared to completers. The magnitude of this lower studying probability and its age‐related trend are almost identical to those observed for young non‐completers dropping out from upper secondary education only after three years, just to withdraw into inactivity. However, when it comes to their employment prospects, the situation looks notably weaker. In particular, also when compared to their non‐completer peers withdrawing only after three years in upper secondary education, their employment prospects are much weaker already when aged 21, with the trend pointing to a faster weakening with age. Also their risk of becoming unemployed is clearly higher. The same pattern is repeated when it comes to their risk of continuing in or returning to inactivity. Moreover, it holds true also in relation to disabil‐ ity benefits, but only for Denmark. In the other three countries, these two groups of early withdrawing non‐completers have an approximately equally high risk, when compared to completers, of ending up in disabil‐ ity arrangements. All in all, the overall impression is that young non‐ completers dropping out from upper secondary education into inactivity


261

fare in most respects worse if dropping out already after two years in‐ stead of dropping out only after three years. The share of non‐completers having dropped out from school into in‐ activity already after having continued in post‐compulsory education for only one year [15555; with 1 standing for full‐time student and 5 for “other” (activities)] is smaller than for the other two groups of young withdrawing non‐completers investigated so far: less than 2% for Fin‐ land and Sweden and 3.5% for Norway. The only exception is Denmark where this share is approximately the same (5%) as for the previous group of withdrawing non‐completers. These young non‐completers’ probabilities in relation to later labour market outcomes, when com‐ pared to the situation of completers, are given in Table 7.11c. Table 7.11c: “Dropout‐after‐one‐year‐withdrawal‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabili‐ ties when compared to completers

Denmark

Finland

Norway

Sweden

[15555] share 5.0%

[15555] share 1.7%

[15555] share 3.5%

[15555] share 1.8%


‐0.209 ‐0.123 0

‐0.238 0 0

‐0.372 ‐0.158 0

‐0.188 0 0


‐0.221 ‐0.219 ‐0.326

‐0.143 ‐0.247 0

0.095 ‐0.074 ‐0.195

‐0.261 ‐0.266 ‐0.295


0.075 0.075 0

0.063 0.126 0

0.087 0.078 0.072

0.107 0.102 0.064


0.060 0.095 0.131

0.039 0.023 ‐0.011

0.032 0.040 0.047

0.129 0.076 0.120


0.294 0.171 0.163

0.279 0.138 0

0.158 0.113 0.097

0.213 0.128 0.074



The broad picture emerging based on Table 7.11c is highly similar to that obtained for young withdrawing non‐completers having been en‐ rolled for two years in upper secondary education before dropping out. There are, however, a few conspicuous differences. First, in all four countries young withdrawing non‐completers dropping out from post‐ compulsory education already after one year, seem to have an even low‐ er probability of having returned to education by age 21. Another dis‐

262


tinct difference compared to young withdrawing non‐completers drop‐ ping out only after two years, is that their probability of continuing in or returning to inactivity is clearly higher. This holds true for Norway and especially for Denmark, that is, for those two countries where the share of non‐completers following this type of inactivity track is highest among the four Nordic countries. Finally, we turn to the pathway indicating immediate withdrawal af‐ ter completion of compulsory education [55555; with 5 standing for “other” (activities)]. The share of non‐completers following this risky track already before turning 21 is less than 2% in Norway, 3.5% in Denmark, but as high as 7% in Sweden and 8% in Finland. Table 7.11d gives the differences in probabilities, when compared to completers, of these young non‐completers of ending up in alternative labour market situations at age 21, 26 and 31, respectively. Table 7.11d: “Early‐withdrawal‐track” non‐completers’ probabilities in terms of labour market outcomes at age 21, 26 and 31, respectively, by country; differences in probabilities when com‐ pared to completers Denmark

Finland

Norway

Sweden

[55555] share 3.5%

[55555] share 7.9%

[55555] share 1.7%

[55555] share 6.9%


‐0.244 ‐0.138 0

‐0.244 ‐0.090 ‐0.051

‐0.426 ‐0.123 0

‐0.254 ‐0.100 ‐0.033


‐0.217 ‐0.254 ‐0.444

‐0.214 ‐0.344 ‐0.343

0.275 ‐0.134 ‐0.185

‐0.320 ‐0.278 ‐0.372


0.082 0.074 0

0.036 0.138 0.109

0.072 0.090 0.068

0.060 0.040 0.046


0.055 0.094 0.137

0.097 0.068 0.088

0 0.047 0.046

0.409 0.246 0.279


0.324 0.224 0.237

0.324 0.228 0.198

0.066 0.120 0.077

0.106 0.092 0.079


Notes: See Table 7.7a above. “5” refers to being in “other” activities.

Young non‐completers withdrawing straight after completing compulso‐ ry school have a very low probability of being enrolled in education at age 21: compared to young completers, this probability is 43% lower in Norway and about 25% lower in the other three countries. Indeed, among the four groups of young non‐completers following inactivity‐


263

dominated tracks early in life, these immediate withdrawers are least likely to have re‐entered education by age 21. They also typically have the weakest employment prospects which, moreover, rapidly weaken further with age: compared to completers, they have in Denmark a 44% lower probability of being employed when aged 31. Their risk of becom‐ ing unemployed is high, but not necessarily higher than for those with‐ drawing non‐completers having continued in education for at least one year before dropping out. When it comes to other types of NEET activities the situation is differ‐ ent, though. In particular, among the four groups of withdrawing non‐ completers, they have the highest probability of moving into disability arrangements, notably in the two countries with the highest share of young non‐completers following this type of track, that is, Finland and Sweden. They also have the highest probability of continuing in or return‐ ing to inactivity, but only in Denmark and Finland, whereas this risk is more in line with that of the other three withdrawing groups in Norway and Sweden. However, despite these cross‐country differences in certain respects, the overall picture points to these immediate withdrawers hav‐ ing the worst prospects when it comes to labour market outcomes in adulthood. On the whole, it seems that these prospects tend to be the weaker, the earlier the young non‐completer withdraws into inactivity.

7.4.6

Main findings

This sub‐chapter has focused on reporting and discussing results high‐ lighting the potential role of young people’s early educational and labour market experiences for their outcomes in adulthood or, more precisely, at age 21, 26 and 31, respectively. These early experiences following upon completion of compulsory school have been approximated by means of 16 stylized pathways constructed for the non‐completers and presented in detail in Chapter 5. With results produced for 16 stylized pathways at three age points and, moreover, separately for four countries, the output from this exer‐ cise is quite massive. While the detailed country‐specific results have been displayed in a large number of tables, the text surrounding these tables has mainly focused on identifying common patterns among the four Nordic countries, but also distinct differences. However, although there do exist differences across the four countries in certain respects, most of these differences seem to relate to the underlying data. More precisely, since the shares of non‐completers going into each of the 16 stylized pathways vary substantially across the four countries, the num‐

264


ber of observations is not always enough for obtaining robust results, especially as we simultaneously account for differences in gender, co‐ hort and family background. When relevant, such uncertainty related to the presented results has been indicated in the text. The probabilities of non‐completers being in alternative labour market situations at age 21, 26 and 31, respectively, depending on the post‐ compulsory‐school track followed up to age 20, are throughout contrasted against the situation of completers, that is, young people having graduated from upper secondary education by age 21. The reported differences in probabilities thus indicate how much more or less likely a non‐completer is to show up in a particular labour market situation later in life, when compared to completers. However, comparisons may also be undertaken in other dimensions based on all these results. In particular, it is also pos‐ sible to compare the outcome across non‐completers following different early tracks. While also such attempts were made already when discuss‐ ing the results displayed in the separate tables, we will conclude this sub‐ chapter with four figures which hopefully shed further light on our results and the conclusions to be drawn based on these findings. The results presented in the subsequent figures are simplified, though, when compared to the detailed tables presented above: the country‐specific probabilities have now been turned into Nordic averag‐ es with the country‐specific shares of non‐completers going into each stylized pathway used as weights. This averaging across countries can be justified in view of the similarity in overall patterns. The use of coun‐ try‐specific pathway shares as weights means, in turn, that major results for each non‐completer pathway is given more weight. This choice is motivated especially in cases where some country’s non‐completers are only weakly represented in the pathway in question. On the whole, then, this averaging exercise produces, for each age point, five results for each of the 16 stylized pathways investigated. For instance, we obtain for the non‐completers’ study‐track pathway [11111] five probabilities showing the average difference with respect to completers at age 21 of being (1) enrolled in education; (2) employed; (3) unemployed; (4) on disability benefits; and (5) in other types of inactivity. Corresponding results are produced for this particular pathway also at age 26 and at age 31. Of course, the same exercise is repeated for the other 15 stylized pathways, implying that we now have for each age point only five labour market outcome probabilities per stylized pathway. We start by presenting these five pathway‐specific points for age 21. This is done in Figure 7.1, but with the early disability‐benefit tracks left out and gathered into a separate figure (Figure 7.4). This choice is guid‐


265

ed by the relatively high probabilities obtained for these tracks: includ‐ ing them with the other stylized pathways, the probabilities of which are of a notably smaller magnitude, would produce a highly unbalanced scatter with most points gathered densely in an unreadable way. Figure 7.1 clearly shows that non‐completers following standard study tracks [11111] are most likely to be enrolled in education also at age 21. Indeed, they have a higher probability of being in education also when compared to 21‐year‐old completers (positively signed difference). For those non‐completers having spent a year in inactivity before continuing in education [51111], the corresponding difference with respect to com‐ pleters is minor, albeit still positive. For all other non‐completers – that is, those having followed other types of early post‐compulsory‐school path‐ ways – the probability of being enrolled in full‐time education when aged 21 is notably lower (negatively signed difference). Figure 7.1: Non‐completers’ probabilities in terms of labour market outcomes at age 21; differences in probabilities, when compared to completers, for 13 stylized pathways as averaged over the four Nordic countries under study

Notes: The figure gives differences in probabilities, when compared to completers, averaged over the four Nordic countries under study. These averages are calculated based on the country‐specific probabilities reported in Tables 7.7a to 7.11d, using the country‐specific shares of non‐completers going into each stylized pathway as weights. The horizontal axis measures these average differences in probabilities with a positive sign indicating a higher probability and a negative sign a lower prob‐ ability, when compared to completers. The vertical axis represents 13 out of the 16 stylized path‐ ways constructed for Nordic non‐completers, as described in Chapter 5. A corresponding presenta‐ tion for the three stylized disability‐benefit pathways is given in Figure 7.4.

266


The probability of being employed at age 21 is highest for non‐ completers having substituted school with work already at an early age. In other words, young non‐completers having started an early and, so it seems, successful employment career appear to experience a clear‐cut employment advantage over both completers and non‐completers hav‐ ing followed other early tracks, still when aged 21. Indeed, non‐ completers having followed other types of early tracks are observed to have the lowest probability not only of being enrolled in education but also of being employed. The probability of experiencing unemployment when 21 years‐of‐age is persistently higher for non‐completers than for completers irrespec‐ tive of the early pathway followed by the non‐completer. The difference in the risk of showing up as an unemployed jobseeker is especially pro‐ nounced for those young non‐completers having early unemployment experiences. A similar pattern is observed for young non‐completers having withdrawn into inactivity already at an early age in the sense that they face the highest risk of being outside both education and the labour market still when aged 21. All in all, Figure 7.1 illustrates well that there is a strong link between young people’s early and later school and labour market experiences. Moreover, these links do not disappear when also accounting for differ‐ ences in family background. Indeed, it is worth emphasising also in this context that the differences in probabilities reported in the figure, as well as in the subsequent figures, are those obtained after taking into account differences in parents’ educational background and wage‐ income level. Hence, while differences in family background can go some way in explaining differences in young people’s labour market outcomes, there is still an independent role played by their early school and labour market experiences. The next two figures illustrate the corresponding situation at age 26 (Figure 7.2) and at age 31 (Figure 7.3). The information contained in these two figures indicates that the strong link between early education‐ al and labour market experiences and the probability of being in alterna‐ tive labour market situations observed at age 21, is clearly discernible also at age 26 and still at age 31.


267

Figure 7.2: Non‐completers’ probabilities in terms of labour market outcomes at age 26; differences in probabilities, when compared to completers, for 13 stylized pathways as averaged over the four Nordic countries under study

Notes: See Figure 7.1 above.

All early non‐completer tracks have, by age 26, resulted in a situation with these young people having a much lower probability of being en‐ rolled as full‐time students, when compared to their peers having al‐ ready by age 21 achieved an upper secondary degree. Noteworthy, how‐ ever, is that the difference in this probability is somewhat smaller for those non‐completers having spent at least three early years in post‐ compulsory education, but without having completed an upper second‐ ary education still when aged 21. Non‐completers with an early em‐ ployment track are more likely to be employed than both completers and other non‐completers still at age 26, but with this employment ad‐ vantage over completers now being notably weaker. The worst employ‐ ment prospects are faced by young people having withdrawn into inac‐ tivity already at an early age. The situation at age 21 indicated that young non‐completers typically face a higher probability, when compared to completers, of being in NEET activities, irrespective of the early track followed. This outcome has not changed by age 26. On the contrary, it seems to have sharpened further with the probability of ending up as a NEET being the higher the fewer years the young person has attended school.

268


Figure 7.3: Non‐completers’ probabilities in terms of labour market outcomes at age 31; differences in probabilities, when compared to completers, for 13 stylized pathways as averaged over the four Nordic countries under study


At age 31, finally, the overall picture looks qualitatively very similar to that observed five years earlier, at age 26. There have occurred certain interesting changes, though. The employment advantage over complet‐ ers of those young people having moved early into working life has, by age 31, turned negative. However, this change in employment prospects does not necessarily mean that the employment situation of these young people has weakened in absolute terms. It rather tells that most young completers of upper secondary school have now finished their further education, entered working life and started a successful career. The strong decline up to age 31 in the share of young people being enrolled as full‐time students also explains the minor differences in student probabilities irrespective of the early track followed. Young people with early experiences from NEET activities have, also when aged 31, the weakest labour market prospects, albeit the difference with respect to completers is no longer as pronounced at age 31 as at age 26. In our last summary figure (Figure 7.4), we illustrate the labour mar‐ ket prospects at age 21, 26 and 31, respectively, of young non‐ completers having followed early tracks dominated by time sent on dis‐ ability benefits. The most conspicuous feature of Figure 7.4 is strong persistence in young people’s disability situation: early school leaving,


269

with the youngster becoming a disability beneficiary already at a young age, most likely results in the youngster being a disability beneficiary also as a young adult. This persistence, however, should not be inter‐ preted as being due to receiving benefits at an early age, but is rather explained by persistence of the cause behind the young person’s early move into disability arrangements. Figure 7.4: Non‐completers’ probability of being a disability beneficiary at age 21, 26 and 31, respectively; differences in probability, when compared to completers, for 3 stylized disability‐dominated pathways as averaged over the four Nordic countries under study


All in all, a major finding is that we observe growing persistence with age among young people having followed early tracks dominated by NEET activities: early unemployment shows up increasingly in unem‐ ployment experiences in adulthood; early disability arrangements tend to continue in adulthood; and early inactivity involves a high risk of con‐ tinuing outside both education and the labour market, a process often ending with a move into disability arrangements. However, also study‐dominated tracks tend to involve a non‐ negligible risk of the non‐completer ultimately ending up in NEET activi‐ ties in adulthood. In view of the results presented in Chapter 6, this out‐ come is likely to concern especially those young people who fail to com‐

270


plete an upper secondary degree also beyond the age of 21, that is, those becoming adult non‐completers. Those who appear to do the best among the non‐completers are the ones moving early, without graduating from upper secondary education, but, nonetheless, successfully into working life. An even better solution also for these young non‐completers could, of course, be an upper sec‐ ondary degree before substituting school with work. This would require attractive graduation routes for youth encountering serious problems in finishing their upper secondary education while having a strong desire to work.


271

8. Summary and discussion Young people follow highly different trajectories from age 16 up to age 20, a time period which is often argued to be the most critical in terms of their future labour market outcomes. In other words, the early post‐ compulsory‐school pathway they happen to follow – for one reason or another – is likely to affect their future options and outcomes in a deci‐ sive way. The focus of this report has been on investigating the look of these early pathways, as well as on exploring their link to labour market outcomes in adulthood. In our analyses, we have traced three full cohorts of youth: those who turned 16 in 1993, 1998 and 2003, respectively. All these young people from four Nordic countries – Denmark, Finland, Norway and Sweden – have been followed up to the year 2008, implying that the longest fol‐ low‐up period covers as many as 15 year. In the Nordic countries, a vast majority of compulsory‐school leavers continues immediately, occasionally only after a break year, in upper secondary education and follows up to age 20, at least, a more or less unbroken track of full‐time‐study years. However, there are also large shares of young people going into other types of early post‐compulsory‐ school tracks, many of which involve risky elements such as prolonged withdrawal outside both education and the labour market. In our anal‐ yses, we have used so‐called clustering techniques to group young peo‐ ple into different clusters depending on the early post‐compulsory‐ school trajectory followed. (Chapter 3) Throughout the report, a distinction is made between young people having completed an upper secondary degree by the time they turn 21 (completers), and young people with an exam only from comprehensive school still at age 21 (non‐completers). Most of our focus has been on the category of young non‐completers. Additionally, attention has been paid to highlighting distinct differences between genders, as well as be‐ tween the three youth cohorts under scrutiny. Within the Nordic countries, Denmark and Sweden can be seen to have polar systems of upper secondary schooling, especially when it comes to vocational training. Denmark has a comprehensive apprentice‐ ship system with the young person gathering large amounts of work experience at employers, whereas Sweden can be characterised as hav‐

ing a typical school system also after the reform in 2011. This difference in upper secondary school systems is clearly reflected in young compul‐ sory‐school‐leavers’ trajectories up to age 20 and, hence, also in their completion rates by age. In Sweden, the typical young person continues for three years in upper secondary schooling, with the non‐completion rate by 21 being, at most, 17% in the three youth cohorts investigated. In Denmark, on the other hand, the corresponding non‐completion rate is as high as 39% in the 1998 cohort (and only slightly lower in the other two cohorts) but with the 21‐year‐old youth being, nonetheless, mostly either in school or in employment. The situation in Norway resembles more that of Denmark than of Sweden, with non‐completion rates occa‐ sionally reaching 32%. Finland, in turn, looks more like Sweden than like Denmark or Norway, with non‐completion rates at age 21 being, at most, close to 20%. (Chapter 2) Even when restricting the analysis of early post‐compulsory‐school trajectories to the group of young non‐completers, a majority is still found to follow standard study tracks, but without completing an upper secondary degree by age 21. Indeed, when clustering young Nordic non‐ completers into groups common to all four countries, 58% of Finnish non‐completers follow a standard or delayed upper secondary study track, with the corresponding share being about 62% for Danish non‐ completers, some 71% for Norwegian non‐completers and close to 76% for Swedish non‐completers. Hence, although these youngsters have not succeeded in achieving an upper secondary degree still by age 21, they have typically spent most of their early post‐compulsory‐school years as full‐time students. Moreover, the share of non‐completers following early study tracks has increased in all four countries over time, that is, when comparing the situation across the three youth cohorts covered by our national datasets. While less than 64% of the Nordic non‐completers from the 1993 cohort followed either standard or delayed study tracks after leaving compulsory school, this share had increased to almost 70% in the 2003 cohort. (Chapter 4) When, as in Chapter 5, we refine our analysis of Nordic non‐ completers by grouping them according to their main activity after three or less years in upper secondary education, full‐time studies still retain their position as also non‐completers’ main type of activity: about 39% of Danish, close to 37% of Norwegian and Swedish and some 35% of Finnish non‐completers go into early post‐compulsory‐school trajecto‐ ries dominated by full‐time studying during the five years up to age 20. Another important type of early post‐compulsory‐school trajectory is characterised by the non‐completer substituting school with work be‐

274


fore graduating from upper secondary school: almost 36% of Danish non‐completers show up in this type of employment‐dominated early track with the corresponding share being close to 31% for Finland, al‐ most 28% for Norway and about 25% for Sweden. However, while we observe a clear‐cut increase across cohorts in the share of young non‐ completers following standard study tracks, there seems to be no com‐ mon Nordic trend for these school‐dropout‐employment tracks: larger shares of Swedish non‐completers from the youngest (2003) cohort follow such employment tracks, whereas the situation is the opposite among Danish non‐completers belonging to different cohorts. All in all, then, the results indicate, once again, that pathways dominated by school and work activities are the most prominent ones also among young non‐completers and that this holds true for all four Nordic coun‐ tries under study. (Chapter 5) The rest of Nordic non‐completers follow early post‐compulsory‐ school‐dropout trajectories ending in so‐called NEET activities, that is, unemployment or inactivity, including disability arrangements. Unem‐ ployment pathways are followed by almost 17% of Swedish non‐ completers, about 12% of Finnish non‐completers, over 9% of Norwe‐ gian non‐completers, but less than 5% of Danish non‐completers. Disa‐ bility benefits are rare among young non‐completers with the share being highest (some 4%) for Sweden and lowest (only 1.3%) for Den‐ mark. In contrast, large shares of young people lacking an upper second‐ ary degree still at age 21 have a history of post‐compulsory‐school tra‐ jectories dominated by other types of inactivity. This share is particular‐ ly high for Norway covering about one‐fourth of the country’s non‐ completers with, moreover, a majority of them having dropped out only after three years in upper secondary school. The corresponding share is just below 20% among Danish and Finnish non‐completers, and about 17% among Swedish non‐completers. (Chapter 5) However, when the share of young non‐completers having followed NEET‐dominated tracks already at an early age is, instead, related to the full youth population, these quite large cross‐country differences in NEET shares become much smaller: about 10% of both Danish and Nor‐ wegian youths are non‐completers going early into NEET‐dominated tracks, with the corresponding share being some 6% for Finland and Sweden. Hence, the much lower completion rate among 21‐year‐olds observed for Denmark and Norway, when compared to Finland and Sweden, is not about young compulsory‐school leavers going into NEET‐ dominated tracks, but more about young people following study‐ or em‐


275

ployment‐dominated tracks without achieving an upper secondary de‐ gree by age 21. (Chapter 5) Next, we explored labour market outcomes at age 21, 26 and 31, re‐ spectively. Of completers aged 21, about 90% were found to be either studying or working. This high “activity” share among young completers is retained, and typically also strengthened, at age 26, with more than 67% of Swedish completers being in employment compared to between 59 and 62% in the other three countries. By age 31, the share of com‐ pleters in employment has increased to over 85% in Denmark and Swe‐ den and to about 76% in Finland and Norway, whereas the share still enrolled in full‐time education was mostly down at about 10%. Con‐ versely, the share of completers in NEET activities is in all four countries quite low, ranging from about 13% in Norway to less than 5% in Den‐ mark. (Chapter 6) The situation looks very different for non‐completers, albeit a strik‐ ingly large proportion of also non‐completers is studying or working in adulthood. At age 31, this share is close to 84% for Danish non‐ completers, compared to about 74% for Swedish non‐completers and some 70% for Finnish and Norwegian non‐completers. However, among the non‐completers we also find large shares of unemployed, disability beneficiaries and those having withdrawn into other types of inactivity. Norway stands out with an extraordinarily large share of non‐ completers being outside both education and the labour market when aged 31, whereas Sweden is characterised by a large share of non‐ completers in disability arrangements when young adults. Finland, in turn, has a comparatively large share of non‐completers showing up as unemployed jobseekers as young adults, which mainly reflects the high unemployment level in the 1990s. (Chapter 6) When these outcomes of young non‐completers at three different age points – 21, 26, and 31, respectively – are, instead, related to the full youth population, the overall cross‐country picture changes, once again, in certain crucial respects. Most notably, while we observe for both Denmark and Norway a comparatively high share of non‐completers among the 21‐year‐olds, many of them are, nonetheless, in employment when young adults. At age 31, about 27% of the Danish youth population investigated cover employed young people who were classified as non‐ completers when aged 21. The corresponding share for Norway is around 19%, but only some 10% for Finland and Sweden. These highly different shares reflect well the cross‐country variation in the number of young people completing an upper secondary degree only beyond age 21 and, in particular, the strikingly different labour market outcomes of

276


these late completers in the four countries. Indeed, in Denmark it does not seem to matter that much if you complete your education by age 21 or 26, whereas it matters a lot if you are a non‐completer still at age 31. In Sweden, in contrast, it matters a lot if you fail to complete an upper secondary degree in the normal time, whereas Finland and Norway fall in‐between these two extremes. Another distinct feature is the conspic‐ uously similar share of the NEET population across the four countries despite remarkable cross‐country differences not least in the share of young people still lacking an upper secondary degree when aged 21. (Chapter 6) Taken together, these findings suggest that cross‐country differences in school‐to‐work‐transition patterns are not necessarily a decisive determinant of the countries’ NEET rates. Accordingly, the dif‐ ferent upper secondary education systems in place should rather be judged by their implications for the development of skills and long‐term productivity, as well as by their costs. There is a huge body of literature providing support for the conten‐ tion that school success and, ultimately, labour market outcomes are closely related to the young person’s family background. Also our results support this broad‐based evidence. It is, therefore, of interest to explore whether the link we observe between early and later school and labour market experiences is, ultimately, a link between family background and later labour market outcomes. Our results strongly suggest that this is not the case: the early post‐compulsory‐school trajectory followed by a young person plays a role also after account has been made for differ‐ ences in family background. In particular, we observe growing so‐called state dependence among non‐completers having followed early tracks dominated by NEET activities: early unemployment shows up in an in‐ creasing risk of experiencing unemployment in adulthood; early disabil‐ ity arrangements tend to continue in adulthood; and early inactivity involves a high risk of continuing outside both education and the labour market also as a young adult, with disability arrangements often being the ultimate outcome. (Chapter 7) However, we do not claim that there is a causal relationship between early school and labour market experi‐ ences and later outcomes. The same underlying cause, such as a disabil‐ ity, may well affect both the pathway through upper secondary educa‐ tion and future labour market prospects. Still, the strong relationship identified between the early trajectory followed and subsequent out‐ comes most likely also includes causal elements implying that, regard‐ less of the underlying cause, evidence‐based information about young people’s pathways that predicts undesirable outcomes may serve as an important source for better targeted policies towards youth.


277

In sum, our report points to four major results: A first major finding is large cross‐country differences in terms of non‐completion rates at different ages with non‐completion from upper secondary education continuing well into the thirties. In Finland and Sweden, the non‐completion rate is comparatively low already among 21‐year‐olds and converges towards 10% when young people are traced up to age 31. In Norway and especially in Denmark, the non‐completion rate among 21‐year‐olds is much higher and has, by age 31, converged towards 20%. Indeed, completion from upper secondary school only beyond age 21 is especially prevalent among young Danes. A second main finding emerging for all four countries is that, alt‐ hough non‐completion predicts risky labour market outcomes in adult‐ hood, employment and studying is by far the dominant activity over the next ten years or so also among those young people who fail to complete an upper secondary degree by age 21. Hence, non‐completion when aged 21 does not automatically imply that the young person is an early school‐leaver or a school dropout having serious difficulties in coping with economic and social life. A third major result is that the different pathways followed by young non‐completers straight after completing compulsory school, are strong‐ ly related to their labour market outcomes in adulthood also after taking into account differences in family background. This implies that also evidence concerning young people’s early school and labour market experiences contains valuable information for policies targeted at youth. Last, but not least, our findings show that the distribution of youth in‐ to study and employment activities versus NEET activities converges in adulthood across the four Nordic countries. This suggests that the merits of different upper secondary education systems should perhaps, first and foremost, be judged by their implications for skills and income for‐ mation rather than by their merits in terms of “producing” NEETs.

278


References Abbott, A. (1983). Sequences of Social Events ‐ Concepts and Methods for the Analy‐ sis of Order in Social Processes, Historical Methods 16, 129–147. http://dx.doi.org/10.1080/01615440.1983.10594107 Abbott, A. & Forrest, J. (1986). Optimal Matching Methods for Historical Sequences, Journal of Interdisciplinary History 16, 471–494. http://dx.doi.org/10.2307/204500 Acemoglu, D. & Autor, D.H. (2011). Skills, Tasks and Technologies: Implications for Employment and Earnings. In: O. Ashenfelter & Card, D. (eds.), Handbook of Labor Economics, Volume 4B. Amsterdam: Elsevier. http://dx.doi.org/10.1016/s0169‐ 7218(11)02410‐5 Albæk, K. (2005). Om Lærepladsproblemet, Nationalokonomisk Tidsskrift 143, 1–25. Albæk, K. (2009). The Danish Apprenticeship System, 1931–2002: The Role of Sub‐ sidies and Institutions, Applied Economics Quarterly 55(1), 39–60. http://dx.doi.org/10.3790/aeq.55.1.39 Albæk, K. (2012). The Incidence of Employment Subsidies for Vocational Training, Applied Economics Quarterly 42, 93–109. http://dx.doi.org/10.3790/aeq.58.2.93 Albæk, K., Asplund, R., Barth, E. & von Simson, K. (2014). Early school leaving and labour market prospects. Chapter 6 in T. Valkonen & Vihriälä, V. (eds), The Nordic model – challenged but capable of reform. Copenhagen: TemaNord 2014:531, 235– 259. http://dx.doi.org/10.6027/TN2014‐531 Asplund, R. (2009). The effect of labour market functioning on employment and unem‐ ployment. Helsinki: Ministry of Employment and the Economy, MEE Publications, Employment and Entrepreneurship 40/2009. (in Finnish with English summary). Asplund, R. & Vanhala, P. (2013). Nuorten työllisyydestä ja työttömyydestä hyvin vaihteleva kuva. Helsinki: ETLA Muistio 19 – 18.12.2013. Asplund, R. & Koistinen, P. (2014). Onko työmarkkinoilla tilaa kaikille? Katsaus eri‐ tyisryhmiin kohdistetun politiikan tuloksiin ja haasteisiin. Helsinki: Työ‐ ja elinkei‐ noministeriön julkaisuja, Työ ja yrittäjyys, 22/2014. Asplund, R. & Vanhala, P. (2014). Peruskoulutodistuksen varassa olevien nuorten polut työelämään – 2000‐luvun alussa peruskoulun päättäneiden kokemuksia. Hel‐ sinki: ETLA Muistio 23, 17.1.2014. Barth, E. & von Simson, K. (2012). Ungdomsarbeidsledighet og konjunkturer, Øko‐ nomiske analyser 5/2012. Becker, G.S. (1964:) Human capital; a theoretical and empirical analysis, with special reference to education. National Bureau of Economic Research General series, Vol no 80. National Bureau of Economic Research; distributed by Columbia University Press, New York. Bell, D.N.F. & Blanchflower, D.G. (2009). What to do about rising unemployment in the UK? Bonn: IZA DP No. 4040.

Björklund, A., Lindahl, L. & Lindquist, M.J. (2010). What More Than Parental Income, Education and Occupation? An Exploration of What Swedish Siblings Get from Their Parents, B e Journal of Economic Analysis & Policy, 10. Black, S. & Devereux, P. (2011). Recent developments in intergenerational mobility. In: O. Ashenfelter & Card, D. (eds.), Handbook of Labor Economics. Amsterdam: Elsevier, 1487–1695. http://dx.doi.org/10.1016/s0169‐7218(11)02414‐2 Bragstad, T. & Brage, S. (2011). Unge på arbeids‐ og helserelaterte ordninger. Oslo: NAV‐rapport 1/2011. Bratsberg, B., Raaum, O., Røed, K. & Gjefsen, H.M. (2010). Utdannings‐ og arbeidskar‐ rierer hos unge voksne: Hvor havner ungdom som slutter skolen i ung alder? Rapport 3/2010, Ragnar Frisch Centre for Economic Research. Brunello. G. & De Paola, M. (2014). The Costs of Early School Leaving in Europe, IZA Journal of Labor Policy 2014, 3:22, 1–11. http://dx.doi.org/10.1186/2193‐9004‐3‐22 Brzinsky‐Fay C. (2007). Lost in Transition? Labour Market Sequences of School‐ leavers in Europe, European Sociological Review 23(4), 409–422. http://dx.doi.org/10.1093/esr/jcm011 Brzinsky‐Fay, C., Kohler, U. & Luniak, M. (2006). Sequence analysis with Stata, Stata Journal 6, 435–460. Bäckman, O., Jakobsen, V., Lorentzen, T., Österbacka, E. & Dahl, E. (2011). Dropping out in Scandinavia. Social Exclusion and Labour Market Attachment among Upper Secondary School Dropouts in Denmark, Finland, Norway and Sweden. Arbetsrapport 2011:8, Institutet för Framtidsstudier. Cedefop (2012). From education to working life: the labour market outcomes of voca‐ tional education and training. http://www.cedefop.europa.eu/EN/publications/21556.aspx Clark, D. (2011). Do Recessions Keep Students in School? The Impact of Youth Un‐ employment on Enrolment in Post‐compulsory Education in England, Economica 78(311), 523–545. http://dx.doi.org/10.1111/j.1468‐0335.2009.00824.x COM (2010) 2020: Europe 2020. A strategy for smart, sustainable and inclusive growth. Brussels: Communication from the Commission of the European Communi‐ ties, 3.3.2010. COM (2010) 477: Youth on the Move – An initiative to unleash the potential of young people to achieve smart, sustainable and inclusive growth in the European Union. Brussels: European Commission, 15.9.2010. COM (2010) 682: An Agenda for new skills and jobs: A European contribution towards full employment. Strasbourg: Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, 23.11.2010. COM (2011) 18: Tackling early school leaving. A key contribution to the Europe 2020 Agenda. Brussels: Communication from the Commission of the European Communi‐ ties, 31.1.2011. COM (2011) 19: Proposal for a Council Recommendation on policies to reduce early school leaving. Brussels, 31.1.2011.

280


COM (2012) 495 final: Draft 2012 Joint Report of the Council and the Commission on the implementation of the renewed framework for European cooperation in the youth field (EU Youth Strategy 2010–2018). Brussels: Communication from the Commis‐ sion to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, 10.9.2012. COM (2011) 933 final: Youth Opportunities Initiative. Brussels: Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, 20.12.2011. Engberg, J. (2014). Rekrytering, genomströmning och relevans – en studie av yrkes‐ och lärlingsutbildningssystemen i Norden. Köpenhamn: Nordiska ministerrådet, TemaNord 2014:544. http://dx.doi.org/10.6027/TN2014‐544 Eurofound (2012). NEETs – Young People not Employment, Education or Training: Characteristics, Costs and Policy Responses in Europe. Dublin. Falch, T. & Nyhus, O.H. (2009). Frafall i videregående opplæring og arbeidsmarked‐ stilknytning for unge voksne. Senter for økonomisk forskning AS. Halvorsen, B., Hansen, O. J. & Tägtström, J. (2012). Unge på kanten. Om inkludering av utsatte ungdommer. København: Nordisk ministerråd, TemaNord 2012:005. http://dx.doi.org/10.6027/Nord2012‐005 Hardoy, I., Røed, K., Torp, H. & Zhang, T. (2006). Virker ungdomsgarantien?, Søkelys på arbeidsmarkedet 23, 21–30. Høst, H. (2008). Fag‐ og yrkesopplæringen i Norge – noen sentrale utviklingstrekk. Oslo: NIFU Step Rapport 20. ILO (2011). Global Employment Trends for Youth: 2011 update. Geneva. ILO (2013). Global Employment Trends 2013: Recovering from a second jobs dip. Geneva. ILO–IMF (2010). The Challenges of Growth, Employment and Social Cohesion. Discus‐ sion document downloadable at http://www.osloconference2010.org/ Jakobsen, V. & Liversage, A. (2010). Køn og etnicitet i uddannelsessystemet litteratur‐ studier og registerdata. København: SFI – Det Nationale Forskningscenter for Vel‐ færd, Rapport nr. 10:29. Kahn, L.B. (2010). The long‐term labor market consequences of graduating from college in a bad economy, Labour Economics 17(2), 303–316. http://dx.doi.org/10.1016/j.labeco.2009.09.002 Larja, L. (2013). Nuorten elinoloja ei voi kuvata pelkän työttömyysasteen avulla, Hy‐ vinvointikatsaus 1/2013, 9–17. Lyche, C.S. (2010). Taking on the Completion Challenge: A Literature Review on Poli‐ cies to Prevent Dropout and Early School Leaving. Paris: OECD Publishing, OECD Ed‐ ucation Working Paper No. 53. http://dx.doi.org/10.1787/5km4m2t59cmr‐en Markussen, E. (Red.) (2010). Frafall i utdanning for 16–20 åringer i Norden. København: Nordisk ministerråd, TemaNord 2010:517. http://dx.doi.org/10.6027/TN2010‐517 Martin, P. & Wiggins, R.D. (2011). Optimal matching analysis. In: Willians, M. & Vogt, W.P. (eds), SAGE Handbook of Innovations in Social Research Methods. Los Angeles: Sage, 385–408. http://dx.doi.org/10.4135/9781446268261.n22 Ministry of Education and Research (2000). Education Act 1985:1100. Stockholm. Myrskylä, P. (2011a). Nuorten työttömyyden mittaaminen on vaikeaa, Tieto&trendit 8/2011.


281

Myrskylä, P. (2011b). Nuoret työmarkkinoiden ja opiskelun ulkopuolella. Helsinki: Työ‐ ja elinkeinoministeriön julkaisuja, Työ ja yrittäjyys 12/2011. Möller, J. & Umkehrer, M. (2014). Are There Long‐Term Earnings Scars from Youth Unemployment in Germany?, ZEW Discussion Paper No. 14‐089. Müller, W. & Gangl, M. (eds.) (2003). Transitions from education to work in Europe. The Integration of Youth into EU Labour Markets. Oxford: Oxford University Press. http://dx.doi.org/10.1093/0199252475.001.0001 Nilsen, Ø. A. & Holm Reiso, K. (2014). Scarring effects of early‐career unemployment, Nordic Economic Policy Review 1/2014, 13–46. Nordström Skans, O. (2011). Scarring Effects of the First Labor Market Experience. Bonn: IZA Discussion Paper No. 5565. OECD (2008a). Employment Outlook 2008. Paris. OECD (2008b). Off to a Good Start? – A Descriptive Review of Youth Labour Market Transitions in OECD Countries. Paris. OECD (2008c). Jobs for Youth: Norway. Paris: OECD Publishing. OECD (2009). Employment Outlook. Paris: OECD Publishing. OECD (2010a). Education at a Glance 2010. OECD INDICATORS. Paris: OECD Publishing. OECD (2010b). Off to a Good Start? Jobs for Youth. Paris: OECD Publishing. OECD (2010c). Jobs for Youth: Denmark. Paris: OECD Publishing. Olofsson, J. & Wadensjö, E. (2007). Ungdomar, utbildning och arbetsmarknaden i Norden – lika men ändå så olika. Stockholm: Rapport från Forskningsrådet för ar‐ betsliv och socialvetenskap. Olofsson, J. & Panican, A. (red.) (2008). Ungdomars väg från skola till arbetsliv – nor‐ diska erfarenheter. Köpenhamn: Nordiska ministerrådet, TemaNord 2008:584. http://dx.doi.org/10.6027/TN2008‐584 Prince Trust (2010). YouGov Youth Index 2010. http://www.princes‐ trust.org.uk/pdf/Youth_Index_2010.pdf Quintini, G. & Manfredi, T. (2009). Going Separate Ways? School‐to‐Work Transitions in the United States and Europe. Paris: OECD Publishing OECD Social, Employment and Migration Working Paper, No. 90. http://dx.doi.org/10.1787/221717700447 Ramböll Management Consulting AB (2010). Nordiska länders insatser mot ung‐ domsarbetslöshet – kartläggning och analys. Köpenhamn: Nordiska ministerrådet, TemaNord 2010:570. http://dx.doi.org/10.6027/TN2010‐570 Reiling, R.B. & Strøm, B. (2015). Upper secondary school completion and the busi‐ ness cycle, The Scandinavian Journal of Economics 117(1), 195–219. Räisänen, H. (2013). Suomessa talouskriisin vaikutuksia on vaimennettu ennen kaik‐ kea lomautuksilla, Hyvinvointikatsaus 1/2013, 24–29. Räisänen, H., Alatalo, J., Krüger Henriksen, K., Israelsson, T. & Klinger, S. (2012). Labour market reforms and performance in Denmark, Germany, Sweden and Finland. Helsinki: Ministry of Employment and the Economy, Employment and Entrepre‐ neurship 19/2012. Saar, E., Unt, M. & Kogan, I. (2008). Transition from Educational System to Labour Market in the European Union: A Comparison between New and Old Members, In‐ ternational Journal of Comparative Sociology 49(1), 31–60. http://dx.doi.org/10.1177/0020715207088586

282


Scarpetta, S., Sonnet, A. & Manfredi, T. (2010). Rising Youth Unemployment During The Crisis: How to Prevent Negative Long‐term Consequences on a Generation? Paris: OECD Social, Employment and Migration Working Papers No. 106. Schmillen, A. & Umkehrer, M. (2013). The Scars of Youth – Effects of Early‐Career Unemployment on Future Unemployment Experiences. IAB Discussion Paper 06/2013. SEC(2011) 96: Reducing early school leaving. Brussels: Commission Staff Working Paper, 31.1.2011. SEC (2011) 401: On EU indicators in the field of youth. Brussels: Commission Staff Working Document, 25.03.2011. Socialstyrelsen (2014). Ekonomiskt bistånd årsstatistik 2013. Stockholm. http://www.socialstyrelsen.se/publikationer2014/2014‐6‐22. Wolbers M. (2007). Patterns of labour market entry: A comparative perspective on school‐to‐work transitions in 11 European countries, Acta Sociologica 50, 189–210. http://dx.doi.org/10.1177/0001699307080924 Von Simson, K. (2014). Explaining upper secondary school dropout: New evidence on the role of local labor markets, Empirical Economics. http://dx.doi.org/10.1007/ s00181‐014‐0829‐3


283

Sammanfattning Unga i åldern 16–20 år gör ofta olika vägval när det gäller utbildning och planering inför sitt framtida yrkesliv. Samtidigt sägs detta vara den tids‐ period i livet som är den mest kritiska med tanke på senare arbetsmark‐ nadsutfall. Med andra ord, de vägar som de unga händelsevis väljer un‐ der de första åren efter avslutad grundskola kommer högst sannolikt att påverka deras framtida möjligheter och utfall på ett avgörande sätt. Det centrala i föreliggande rapport har varit att undersöka hur dessa tidiga vägval ser ut, samt att utforska deras koppling till de ungas arbetsmark‐ nadsutfall i vuxenlivet. I vår studie följer vi tre kohorter av ungdomar: de som fyllde 16 år 1993, 1998 och 2003. Alla dessa ungdomar, som representerar totalt fyra nordiska länder – Danmark, Finland, Norge och Sverige, har följts fram till år 2008, vilket innebär att den längsta uppföljningsperioden omfattar 15 år. I de nordiska länderna inleder en stor majoritet grundskoleelever fortsatta studier, antingen direkt eller efter ett mellanår, och studerar mer eller mindre oavbrutet fram till 20 års ålder, vanligtvis ännu längre. Många unga gör emellertid också en helt andra vägval, vilka ofta kantas av riskfyllda faser såsom längre perioder utanför både utbildning och arbetsmarknad. I våra analyser har vi använt en så kallad klusterteknik för att gruppera de unga i olika kluster beroende på vilka vägval de gjort efter avslutad grundskola. Genom hela rapporten görs konsekvent en distinktion mellan unga som har fullföljt sina fortsatta studier efter avslutad grundskola senast vid fyllda 21 år, och unga som fortfarande endast har grundskoleexamen som 21‐åringar.7 Tyngdpunkten i rapporten ligger på den senare kate‐

────────────────────────── 7 I vår rapport avser fortsatta studier efter avslutad grundskola det som på engelska kallas upper secondary

education och som motsvarar ISCED 3. Generellt kan denna utbildningsnivå betecknas som utbildning på mellannivå efter avslutad utbildning på lägre och högre grundnivå (= grundskola). Denna utbildning på mellannivå är emellertid olika utformad också i de nordiska länderna och beskrivs dessutom med olika mer eller mindre officiella termer. I Danmark används numera termen ”ungdomsuddannelse”. I Finland talar man om ”fortsatta studier efter grundskolan i den allmänbildande gymnasieutbildningen och yrkesutbildningen”, medan man inom utbildningsförvaltningen talar om andra stadiet i motsats till grundnivå (=ISCED 1&2). I

gorin av unga. Dessutom lyfter vi fram skillnader mellan könen, liksom mellan de tre undersökta ungdomskohorterna. När det gäller de nordiska länderna kan Danmark och Sverige närm‐ ast beskrivas som varandras motpoler i fråga om den fortsatta utbild‐ ningen efter avslutad grundskola, särskilt yrkesutbildningen. Danmark har ett välutvecklat system för lärlingsutbildning som ger de unga om‐ fattande arbetsplatsförlagd praktik, medan yrkesutbildningen i Sverige i huvudsak kan beskrivas som skolbaserad, även efter reformen som trädde i kraft 2011. Denna skillnad i ländernas utbildningssystem efter avslutad grundskola avspeglas tydligt i de ungas levnadsbanor fram till 20 års ålder och därför även i andelen unga som vid en given ålder full‐ följt sina studier efter avslutad grundskola. I Sverige fortsätter den ty‐ piska unga att studera i tre år efter avslutad grundskola och andelen unga som inte fullföljt sina fortsatta studier som 21‐åringar är som högst 17 % i de tre ungdomskohorterna. I Danmark, å andra sidan, är motsva‐ rande andel unga så hög som 39 % i 1998 års ungdomskohort (och end‐ ast något lägre i de andra två kohorterna). Situationen i Norge liknar mer den i Danmark än i Sverige, med en andel unga på som högst 32 % som har enbart grundskoleexamen fortfarande vid 21 års ålder. Finland ser mer ut som Sverige än som Danmark eller Norge, med en andel på knappt 20 % som inte ännu vid 21 års ålder avlagt någon examen efter avslutad grundskola. Det är anmärkningsvärt att även när vi begränsar analysen till de unga som fortfarande som 21‐åringar inte avlagt någon examen efter avslutad grundskola så visar det sig att en betydande del av dem fortsatt studera, men utan att lyckas slutföra sina studier inom fem år efter av‐ slutad grundskola. När vi gör en gemensam nordisk klusteranalys för dessa ungdomar finner vi att 58 % av de finländska ungdomarna har fortsatt sina studier efter avslutad grundskola, antingen direkt eller efter ett mellanår, men utan att ha avlagt examen som 21‐åringar. Motsva‐ rande andel för Danmark är ca 62 %, för Norge ungefär 71 % och för Sverige nästan 76 %. Dessa andelar vittnar om att även om de här ung‐ domarna inte lyckats fullfölja sina fortsatta studier ännu vid 21 års ålder så har de oftast tillbringat största delen av tiden efter avslutad grund‐ skola i utbildning. Den här andelen unga har dessutom ökat över tid i samtliga fyra länder, vilket framgår när man jämför situationen i de tre ungdomskohorter som vår studie bygger på. Knappt 64 % av de nor‐ diska unga i 1993 års kohort som fortfarande som 21‐åringar inte avlagt

Norge används termen ”videregående avsluttende utdanning” medan man i Sverige använder begreppet ”utbildning på gymnasial nivå”.

286


någon examen efter avslutad grundskola hade fortsatt studera efter grundskolan antingen direkt eller efter ett mellanår. Andelen hade ökat till nästan 70 % i 2003 års kohort. En något mer detaljerad analys av dessa ungdomar utgående från de‐ ras huvudsakliga verksamhet efter tre eller färre år i fortsatta studier efter avslutad grundskola, utvisar att fortsatta studier är den övervä‐ gande vanligaste aktiviteten: i Danmark är ca 39 %, i Norge och Sverige nästan 37 % och i Finland omkring 35 % av dessa unga huvudsakligen i utbildning under de fem första åren efter avslutad grundskola. En annan viktig grupp bildar de unga som börjat arbeta istället för att fullfölja sina fortsatta studier efter avslutad grundskola: av de danska ungdomar som fortfarande som 21‐åringar saknar examen från fortsatta studier dyker nästan 36 % upp i den här jobbdominerade kategorin, medan motsva‐ rande andel är knappt 31 % för Finland, närmare 28 % för Norge och ca 25 % för Sverige. Men medan vi ser en tydlig ökning över tid av andelen unga som fortsätter i utbildning utan att avlägga examen, så förefaller det inte finnas någon gemensam nordisk trend för de unga som avbrutit sina fortsatta studier för arbete: i Sverige visar andelen unga med enbart grundskoleexamen som börjat jobba en ökning över tid, medan situat‐ ionen är den motsatta bland de danska årskullarna. Överlag tyder resul‐ taten på att ungas vägval efter avslutad grundskola domineras av an‐ tingen fortsatta studier eller jobb eller en kombination av dessa. De här vägvalen är de mest framträdande även bland de ungdomar som fortfa‐ rande som 21‐åringar saknar examen efter avslutad grundskola. Dessu‐ tom upprepas samma mönster i samtliga fyra nordiska länder som ingår i vår studie. Resten av de unga som fortfarande vid 21 års ålder saknar examen ef‐ ter avslutad grundskola tenderar att höra till den så kallade NEET‐ gruppen, vilket innebär att de antingen är arbetslösa eller befinner sig helt utanför arbetskraften (= inaktiva, inklusive unga med funktionshin‐ der). Andelen 21‐åringar med enbart grundskoleexamen som mestadels varit arbetslösa under de fem första åren efter avslutad grundskola är nära 17 % i Sverige, medan motsvarande andel är ca 12 % i Finland, drygt 9 % i Norge men under 5 % i Danmark. Förtidspensioneringar fö‐ rekommer sällan; andelen är högst (ca 4 %) för Sverige och lägst (endast 1,3 %) för Danmark. Däremot är andra former av inaktivitet mycket van‐ ligt förekommande bland unga som fortfarande som 21‐åringar saknar examen efter avslutad grundskola. Andelen är särskilt hög för Norge eller ca 25 %. För Danmark och Finland stannar andelen strax under 20 %, och för Sverige är den ca 17 %.


287

Dessa relativt stora skillnader mellan de nordiska ländernas NEET‐ andelar minskar dock märkbart om andelen unga som huvudsakligen följt NEET‐dominerade vägar efter avslutad grundskola relateras till hela ungdomspopulationen och inte, som ovan, till populationen av unga som fortfarande som 21‐åringar saknar examen efter avslutad grund‐ skola: ca 10 % av både danska och norska ungdomar kan karakteriseras som unga som har enbart grundskoleexamen på grund av att de i ett tidigt skede efter avslutad grundskola hamnat utanför både utbildning och arbetsliv. Motsvarande andel för Finland och Sverige är ca 6 %. Detta innebär att den betydligt lägre andelen danska och norska unga (jämfört med finländska och svenska unga) som fullföljt sina fortsatta studier senast vid 21 års ålder, inte bör tolkas som att danska och norska unga uppvisar en klart högre risk att hamna i en NEET‐situation efter avslutad grundskola utan snarare som att de har en större tendens att fortsätta i utbildning eller börja jobba utan att ännu som 21‐åringar ha avlagt examen efter avslutad grundskola. Slutligen undersökte vi de ungas arbetsmarknadsutfall vid 21, 26 re‐ spektive 31 års ålder. Av 21‐åringar med examen efter avslutad grund‐ skola var ca 90 % i antingen utbildning eller arbete. Samma eller ännu större andelar noterades för de högre åldrarna. Vid 26 års ålder var mer än 67 % av de svenska ungdomar som avlagt examen efter avslutad grundskola senast som 21‐åringar, sysselsatta jämfört med mellan 59 och 62 % i de tre övriga länderna. Vid 31 års ålder hade andelen syssel‐ satta bland dessa unga ökat till över 85 % i Danmark och Sverige och till ca 76 % i Finland och Norge, medan andelen som fortfarande var i ut‐ bildning hade vanligtvis sjunkit till runt 10 %. Omvänt var NEET‐ andelen bland dessa unga i alla fyra länder relativt låg, ca 13 % i Norge och under 5 % i Danmark. Situationen ser helt annorlunda ut för de unga som ännu som 21‐ åringar saknade examen efter avslutad grundskola, även om en påfal‐ lande stor del av också dessa ungdomar antingen studerar eller arbetar i vuxen ålder. Vid 31 års ålder är den här andelen närmare 84 % för de danska ungdomar som fortfarande vid 21 års ålder hade enbart grund‐ skoleexamen, medan motsvarande andel är ca 74 % för Sverige och om‐ kring 70 % för Finland och Norge. Men i den här specifika ungdoms‐ gruppen finner vi också många som vid 31 års ålder är arbetslösa, erhål‐ ler ersättning kopplat till funktionshinder eller befinner sig av andra orsaker utanför både utbildning och arbetsmarknad. Norge sticker ut med en mycket stor andel unga utanför utbildning och arbetsliv, medan Sverige kännetecknas av en stor andel som erhåller någon form av er‐ sättning kopplat till funktionshinder. Finland åter har bland dessa unga

288


en jämförelsevis stor andel som är arbetslösa som unga vuxna, något som i huvudsak speglar den höga arbetslösheten under 1990‐talet och särskilt bland dem med enbart grundskoleexamen. När vi istället relaterar dessa arbetsmarknadsutfall för unga som fortfarande som 21‐åringar saknar examen efter avslutad grundskola till hela ungdomspopulationen, förändras dock återigen skillnaderna mellan länderna. Mest anmärkningsvärt är att trots den höga andelen danska och norska ungdomar som fortfarande vid 21 års ålder saknar examen efter avslutad grundskola så är många av dem de facto sysselsatta som unga vuxna. Vid 31 års ålder representerar ca 27 % av den danska ung‐ domspopulationen som täcks av vår studie sysselsatta unga som vid 21 års ålder hade enbart grundskoleexamen. Motsvarande andel för Norge är omkring 19 %, men endast ca 10 % för Finland och Sverige. Den här stora variationen i andelar avspeglar väl det faktum att länderna uppvi‐ sar väsentliga skillnader i antalet unga som fullföljer sina fortsatta stu‐ dier först efter att de fyllt 21, och i all synnerhet att arbetsmarknadsut‐ fallet för dessa unga är högst olika i de fyra länderna. I Danmark verkar det faktiskt inte spela så stor roll om man fullföljer sina fortsatta studier som 21‐ eller 26‐åring, medan det däremot betyder en hel del om du fortfarande som 31‐åring saknar examen efter avslutad grundskola. I Sverige, å andra sidan, är det av avgörande betydelse för arbetsmark‐ nadsutfallet om du misslyckas med att fullfölja dina fortsatta studier inom normal tid. Finland och Norge placerar sig mellan dessa två ex‐ tremfall. Ett annat anmärkningsvärt resultat är att NEET‐andelen i ung‐ domspopulationen är i stort sett densamma i de fyra länderna trots de stora skillnaderna i andelen unga som fortfarande vid 21 års ålder sak‐ nar examen efter avslutad grundskola. Sammantaget tyder dessa resul‐ tat på att skillnader mellan länder i fråga om de ungas övergång från skola till arbetsliv inte nödvändigtvis är av avgörande betydelse för län‐ dernas NEET‐grad. Detta tyder i sin tur på att de olika systemen för fort‐ satta studier efter avslutad grundskola bör snarare utvärderas på basis av deras effekt på kunskapsutvecklingen och den långsiktiga produktivi‐ teten, samt de kostnader de ger upphov till. Det finns en omfattande litteratur som stöder antagandet att skolre‐ sultat och i sista hand arbetsmarknadsutfall är nära relaterade till den ungas familjebakgrund. Också våra resultat stöder detta antagande. Följaktligen är det av intresse att undersöka om sambandet vi ser mellan tidiga och senare utbildnings‐ och arbetsmarknadserfarenheter egentli‐ gen är en koppling mellan familjebakgrund och senare arbetsmarknads‐ utfall. Våra resultat tyder på att så absolut inte är fallet: de vägval som ungdomar gör efter avslutad grundskola är av stor betydelse också efter


289

att vi kontrollerat för väsentliga skillnader i de ungas familjebakgrund. Framför allt ser vi tilltagande så kallat situationsberoende (eng. state‐ dependence) bland unga som i ett tidigt skede efter avslutad grundskola hamnat i en situation som domineras av NEET‐aktiviteter. Tidiga erfa‐ renheter av arbetslöshet visar sig i form av en klart ökad risk att bli ar‐ betslös som ung vuxen. Ersättningar kopplade till funktionshinder ten‐ derar att fortsätta in i vuxenlivet. Tidiga erfarenheter av andra former av inaktivitet medför ökad risk att den unga stannar utanför både utbild‐ ning och arbetsmarknad också som ung vuxen, varvid ersättning till följd av funktionshinder ofta blir den slutliga lösningen. Därmed hävdar vi emellertid inte att det förekommer ett kausalt samband mellan tidiga utbildnings‐ och arbetsmarknadserfarenheter och senare utfall. Samma underliggande orsak, såsom exempelvis funktionshinder, kan mer än väl påverka både de ungas väg genom utbildningssystemet och deras fram‐ tida arbetsmarknadsutsikter. Likväl är det högst sannolikt att det starka samband som vi finner mellan de ungas tidiga vägval och senare utfall samtidigt avspeglar kausala element. Detta innebär i sin tur att, obero‐ ende av den underliggande orsaken, så kan empiriskt framtagen inform‐ ation om ungas vägval som tenderar att leda till oönskade utfall använ‐ das som viktigt bakgrundsmaterial då man eftersträvar att föra en mer målinriktad ungdomspolitik. Sålunda kan vår rapport sammanfattas i form av fyra huvudresultat: Det första huvudresultatet pekar på stora skillnader mellan de fyra studerade nordiska länderna vad gäller unga som saknar examen efter avslutad grundskola; skillnaderna är stora vid olika åldrar och kvarstår också bland unga vuxna. I Finland och Sverige är andelen unga med en‐ bart grundskoleexamen relativt liten redan bland 21‐åringar och om‐ kring 10 % bland 31‐åringar. I Norge, och i synnerhet i Danmark, är an‐ delen 21‐åringar utan examen efter avslutad grundskola betydligt större även om den sjunker till omkring 20 % för 31‐åringar. Det att man avslu‐ tar sina fortsatta studier först efter fyllda 21 är särskilt vanligt bland unga i Danmark. En andra huvudsaklig slutsats som därtill gäller samtliga fyra länder är att även om avsaknaden av examen efter avslutad grundskola ökar risken för sämre arbetsmarknadsutfall i vuxen ålder, så är sysselsättning och utbildning de helt dominerande aktiviteterna ända upp i trettioårs‐ åldern också bland de ungdomar som har enbart grundskoleexamen fortfarande vid 21 års ålder. Enbart grundskoleexamen ännu som 21‐ åring innebär sålunda inte att den unga kan automatiskt klassas som en person som i ett tidigt skede lämnat all utbildning bakom sig eller som

290


en person som avbrutit sina studier till följd av allvarliga ekonomiska och sociala problem. Ett tredje viktigt resultat är att de tidiga vägval som de ungdomar gjort som fortfarande vid 21 års ålder saknar examen efter avslutad grundskola är starkt relaterade till deras arbetsmarknadsutfall som unga vuxna också efter att vi beaktat skillnader i ungdomarnas familje‐ bakgrund. Detta innebär att empiriska resultat gällande ungdomars ti‐ diga utbildnings‐ och arbetsmarknadserfarenheter kan bidra med värde‐ full information när det gäller att utforma åtgärder riktade mot unga. Slutligen utvisar våra resultat att ungdomarna i de fyra studerade länderna fördelar sig rätt olika mellan dels studier och sysselsättning och dels NEET‐aktiviteter, men också att dessa skillnader minskar i be‐ tydande utsträckning när ungdomarna når vuxen ålder. Detta antyder att fördelarna med olika utbildningssystem för fortsatta studier efter avslutad grundskola bör kanske i första hand utvärderas utgående från hur de påverkar kunskaps‐ och inkomstbildningen snarare än i vilken utsträckning de ”producerar” NEET‐unga.


291

Appendix: Descriptions of national datasets used Denmark The Danish dataset used in the project is compiled from register data at Statistics Denmark. The dataset encompasses three cohorts of young Danish residents. The first cohort contains all residents, who were 16 years old in 1993, and they are followed until they are 31 years old, in 2008. The second cohort consists of Danish residents, who were 16 years old in 1998. They are followed up to age 26, in 2008. The third cohort consists of Danish residents, who were 16 years old in 2003, and they are followed until they are 21 years old, in 2008. Statistics Denmark has supplied various kind of information from registers in Statistics Denmark for the young people in these three co‐ horts. The result is linked data constructed for use in the analyses pre‐ sented in this report. The main information used for the project is register‐based labour force statistics (Registerbaseret Arbejdsstyrkestatistik – RAS) that as‐ sign each Danish resident to one and only one state in the Danish labour market at one particular point in time (the end of November each year). This statistic follows the international guidelines of the International Labour Organization (ILO) for assessments of the attachment of the population to the labour market. The ILO guidelines are primarily used in relation to interviews, where each person answers questions about the attachment to the labour market. RAS uses the guidelines to choose the appropriate labour market state. The Danish classification is rather detailed and presently contains 62 different categories, see http://www.dst.dk/da/Statistik/dokumentation/Times/ida‐databasen/ ida‐personer/pstill.aspx. Another main source of information is statistics on education. These statistics contain the current and finished education on a detailed basis. We have also linked the young people in the three cohorts to infor‐ mation on parental background. This background includes the education and the income of the parents (as obtained from tax registers).

Finland The project uses the linked employer‐employee data (Finnish Longitu‐ dinal Employer‐Employee Data FLEED) that Statistics Finland has creat‐ ed for research use. The FLEED data consists of persons aged 15 to 70 living in Finland between 1988 and 2011 (excl. Åland). The persons have been followed over time, so there is data on the person for all the years during which the person has been alive, aged between 15 and 70 and residing in Finland. The FLEED data includes information on the per‐ son’s basic characteristics, family, living, employment relationships, periods of unemployment, income and education. The FLEED data is based on employment statistics that are annual statistics providing data by region on the population’s economic activity and employment. The population for the statistics is the permanently resident population in the country on the last day of the year. The data are mainly derived from around 40 administrative registers and statisti‐ cal data files. The produced data describe the population’s main type of activity, oc‐ cupation, status in occupation, number of workplaces, location of work‐ place, and education and income of the population and labour force. The reference period of the statistics is the last week of the year, but the statis‐ tics also contain data accumulated during the statistical reference year (e.g. income data, months of employment and unemployment). The population and definitions of the employment statistics have re‐ mained more or less the same since 1987, when regular production of the statistics was started. However, the classifications used in the statistics have changed along the years. For example, the Standard Industrial Classi‐ fication was amended in 1993, 2001 and 2007 and the Classification of Occupations in 1995 and 2010. The changes in the classifications have an effect on the comparability of the years because it has not been possible to build complete conversion keys between all the classifications. Sources  http://tilastokeskus.fi/tup/mikroaineistot/ me_kuvaus_henkilo_en.pdf  http://www.stat.fi/meta/til/tyokay_en.html

294


Norway The Norwegian analyses use data from FD‐Trygd, an administrative da‐ tabase compiled by Statistics Norway. The database consists of several welfare and employment registers at the individual level, together with demographic information such as age, gender, parents’ education and income, etc. Information about education is gathered from NUDB (Nas‐ jonal Utdanningsdatabase). The different registers are linked together by a unique identification key, which makes it possible to follow individ‐ uals as they “move” between different labour market and welfare states, as well as to and from education. The data used in this project encompasses all Norwegian residents who have started an upper secondary education during the period 1992–2008. This means that we lack information about youth who do not enrol in upper secondary education in this time period. The enrol‐ ment rate in Norway is very high – around 96–97% of each youth cohort enrol in upper secondary education the same year they finish compulso‐ ry school. Among those who do not have a direct transition from com‐ pulsory to upper secondary school, almost 70% have enrolled within five years after completing compulsory school (Falch and Nyhus, 2009). This means that the data covers practically the entire population of Norwegian youth, but the bias is likely to be larger for the younger co‐ horts as they are observed for fewer years. Table A1 shows the bias for the three youth cohorts used in this project. Table A1. Youth cohorts in the data and full population, Norway Cohort of 16‐year‐olds 1993 1998 2003

Number of youth in the project data

Number of youth in full population

Difference between data and full population

51,012 51,394 53,758

51,282 52,029 55,524

‐270 ‐635 ‐1,766

Notes: The full population numbers are gathered from Statistics Norway. All numbers comprise youth who turn 16 in 1993, 1998 or 2003 and who are resident in Norway the following five years.

The following registers are used in order to classify the youth into the five main activity categories: Ongoing education (Student); Employ‐ ment (Employed); Maternity benefit and sickness benefit (Em‐ ployed); Conscripts (Employed); Job seekers and occupationally handicapped (Unemployed); Rehabilitation and vocational rehabilita‐ tion allowance (Pensioner); Temporary disability benefits (Pension‐ er); Disability pension (Pensioner); Social assistance (Other). In or‐ der to make the data comparable to the other countries’ data in this


295

project, we use December as our reference month. The exception is ongoing education, which is measured in October each year.

Sweden The Swedish data is based on several registers from Statistics Sweden. The major data source is LISA (Longitudinal integration database for health insurance and labour market studies), a longitudinal database covering education, income and employment. The population consists of all over 16‐year‐old nationally registered individuals during the years 1960–2012. This means that a large part of the Swedish population is followed over a long period of time. Information on the index persons’ biological siblings, half siblings on each parent’s side, spouses and all persons residing in the household is added to the database. The different registers are linked together by a unique identification key, which makes it possible to follow individuals as they move between different activi‐ ties. The material also covers variables representing the individuals’ demographic and socio‐economic status. These additional variables make it possible to estimate the effects of failing or succeeding with var‐ ious upper secondary educations, taking into account the individuals’ sex, country of birth, parents’ level of education, parents’ level of income, etc. Thus, the effects of labour market change on young individuals’ la‐ bour market careers can be pursued not only from a macro‐economic perspective, but also from a detailed, geographical and longitudinal so‐ cio‐economic perspective. The Swedish data used in this project include the following main reg‐ isters: Longitudinal integration database for health insurance and labour market studies; the multi‐generational register; the income and wealth register, the education register and the total population register.

296


Table A2. Summary of the number of young people in the four national datasets used Total

Denmark Finland Norway Sweden

Completers


Non‐completers


All three cohorts

Cohort of 16‐year‐olds

All

Males

Females

164,879 193,567 156,164 290,257

85,190 99,211 79,924 149,713

79,689 94,356 76,240 140,544

All three cohorts4

1993

1998

2003

56,710 65,595 51,012 90,611

51,055 67,068 51,394 92,495

57,114 60,904 53,758 107,151


All

Males

Females

103,203 158,611 109,723 243,763

49,046 78,474 53,000 123,438

54,157 80,137 56,723 120,325

All three cohorts

1993

1998

2003

36,932 55,088 36,451 77,498

31,071 53,847 36,434 76,790

35,200 49,676 36,838 89,475


All

Males

Females

61,676 34,956 46,441 46,494

36,144 20,737 26,924 26,275

25,532 14,219 19,517 20,219

1993

1998

2003

19,778 10,507 14,561 13,113

19,984 13,221 14,960 15,705

21,914 11,228 16,920 17,676


297

TemaNord 2015:548


TemaNord 2015:548

Ved Stranden 18 DK-1061 Copenhagen K www.norden.org

Youth unemployment and inactivity A comparison of school-to-work transitions and labour market outcomes in four Nordic countries

Young people follow highly different trajectories from age 16 up to age 20, a time period which is often argued to be the most critical in terms of their future labour market outcomes. The focus of this report is on investigating the look of these early pathways, as well as on exploring their link to labour market outcomes in adulthood. Results are reported and compared for four Nordic countries: Denmark, Finland, Norway and Sweden.

TemaNord 2015:548 ISBN 978-92-893-4229-2 (PRINT) ISBN 978-92-893-4230-8 (PDF) ISBN 978-92-893-4231-5 (EPUB) ISSN 0908-6692

TN2015548 omslag.indd 1

20-07-2015 12:48:57